MIT White Paper on Convergence

合集下载

OpenFlow-Enabling-Innovation-in-Campus-Network以及中文翻译

附录A 外文原文OpenFlow: Enabling Innovation in Campus NetworksNick McKeown Stanford University Guru Parulkar Stanford UniversityTom AndersonUniversity of WashingtonLarry PetersonPrinceton UniversityHari BalakrishnanMITJennifer RexfordPrinceton UniversityScott Shenker University of California,BerkeleyJonathan Turner Washington University inSt. LouisThis article is an editorial note submitted to CCR. It has NOT been peer reviewed.Authors take full responsibility for this article’s technical ments can be posted through CCR Online.ABSTRACTThis whitepaper proposes OpenFlow: a way for researchers to run experimental protocols in the networks they use every day. OpenFlow is based on an Ethernet switch, with an internal ﬂow-table, and a standardized interface to add and remove ﬂow entries. Our goal is to encourage networking vendors to add OpenFlow to their switch products for deployment in college campus backbones and wiring closets. We believe that OpenFlow is a pragmatic compromise: on one hand, it allows researchers to run experiments on heterogeneous switches in a uniform way at line-rate and with high port-density; while on the other hand, vendors do not need to expose the internal workings of their switches. In addition to allowing researchers to evaluate their ideas in real-world traffic settings, OpenFlow could serve as a useful campus component in proposed large-scale testbeds like GENI. Two buildingsat Stanford University will soon run OpenFlow networks, using commercial Ethernet switches and routers. We will work to encourage deployment at other schools; and we encourage you to consider deploying OpenFlow in your university network too.Categories and Subject DescriptorsC.2 [Internetworking]: RoutersGeneral TermsExperimentation, DesignKeywordsEthernet switch, virtualization, flow-based1. THE NEED FOR PROGRAMMABLE NETWORKSNetworks have become part of the critical infrastructure of our businesses, homes and schools. This success has been both a blessing and a curse for networking researchers; their work is more relevant, but their chance of making an impact is more remote. The reduction in real-world impact of any given network innovation is because the enormous installed base of equipment and protocols, and the reluctance to experiment with production traffic, which have created an exceedingly high barrier to entry for new ideas. Today, there is almost no practical way to experiment with new network protocols (e.g., new routing protocols, or alternatives to IP) in sufficiently realistic settings (e.g., at scale carrying real traffi c) to gain the conﬁdence needed for their widespread deployment. The result is that most new ideas from the networking research community go untried and untested; hence the commonly held belief that the network infrastructure has “ossiﬁed”.Having recognized the problem, the networking community is hard at work developing programmable networks, such as GENI [1] a proposed nationwide research facility for experimenting with new network architectures and distributed systems. These programmable networks call for programmable switches and routers that (using virtualization) can process packets for multiple isolated experimental networks simultaneously. For example, in GENI it is envisaged that a researcher will be allocated a slice of resources across the whole network, consisting of a portion of network links, packet processing elements (e.g. routers) and end-hosts; researchers program their slices to behave as they wish. A slice could extend across the backbone, into access networks, into college campuses, industrial research labs, and include wiring closets, wireless networks, and sensor networks.Virtualized programmable networks could lower the barrier to entry for new ideas, increasing the rate of innovation in the network infrastructure. But the plans for nationwide facilities are ambitious (and costly), and it will take years for them to be deployed.This whitepaper focuses on a shorter-term question closer to home: As researchers, how can we run experiments in our campus networks? If we can ﬁgure out how, we can start soon and extend the technique to other campuses to beneﬁt the whole community.To meet this challenge, several questions need answering, including: In the early days, how will college network administrators get comfortable putting experimental equipment (switches, routers, access points, etc.) into their network? How will researchers control a portion of their local network in a way that does not disrupt others who depend on it? And exactly whatfunctionality is needed in network switches to enable experiments? Our goal here is to propose a new switch feature that can help extend programmability into the wiring closet of college campuses.One approach -that we do not take -is to persuade commercial “name-brand” equipment vendor s to provide an open, programmable, virtualized platform on their switches and routers so that researchers can deploy new protocols, while network administrators can take comfort that the equipment is well supported. This outcome is very unlikely in the short-term. Commercial switches and routers do not typically provide an open software platform, let alone provide a means to virtualize either their hardware or software. The practice of commercial networking is that the standardized external interfaces are narrow (i.e., just packet forward ing), and all of the switch’s internalﬂexibility is hidden. The internals di ffer from vendor to vendor, with no standard platform for researchers to experiment with new ideas. Further, network equipment vendors are understandably nervous about opening up interfaces inside their boxes: they have spent years deploying and tuning fragile distributed protocols and algorithms, and they fear that new experiments will bring networks crashing down. And, of course, open platforms lower the barrier-to-entry for new competitors.A few open software platforms already exist, but do not have the performance or port-density we need. The simplest example is a PC with several network interfaces and an operating system. All well-known operating systems support routing of packets between interfaces, and open-source implementations of routing protocols exist (e.g., as part of the Linux distribution, or from XORP [2]); and in most cases it is possible to modify theoperating system to process packets in almost any manner (e.g., using Click [3]). The problem, of course, is performance: A PC can neither support the number of ports needed for a college wiring closet (a fan out of 100+ ports is needed per box), nor the packet-processing performance (wiring closet switches process over 100Gbits/s of data, whereas a typical PC struggles to exceed 1Gbit/s; and the gap between the two is widening).Existing platforms with specialized hardware for line-rate processing are not quite suitable for college wiring closets either. For example, an ATCA-based virtualized programmable router called the Supercharged Planet Lab Platform [4] is under development at Washington University, and can use network processors to process packets from many interfaces simultaneously at line-rate. This approach is promising in the long-term, but for the time being is targeted at large switching centers and is too expensive for widespread deployment in college wiring closets. At the other extreme is NetFPGA [5] targeted for use in teaching and research labs. NetFPGA is a low-cost PCI card with a user-programmable FPGA for processing packets, and 4 ports of Gigabit Ethernet. NetFPGA is limited to just four network interfaces—insufficient for use in a wiring closet.Thus, the commercial sol utions are too closed and inﬂex ible and the research solutions either have insufficient performance or fan out, or are too expensive. It seems unlikely that the research solutions, with their complete generality, can overcome their performance or cost limitations. A more promising approach is to compromise on generality and to seek a degree of switch flexibility that is:Figure 1 Idealized OpenFlow Switch. The Flow Table is controlled by a remote controllervia the Secure Channel.•Amenable to high-performance and low-cost implementations.•Capable of supporting a broad range of research.•Assured to isolate experimental traffic from production traffic. •Consistent with vendors’ need for closed platforms.This paper describes the OpenFlow Switch—a specification that is an initial attempt to meet these four goals.2. THE OPENFLOW SWITCHThe basic idea is simple: we exploit the fact that most modern Ethernet switches and routers contain ﬂow-tables (typically built from TCAMs) that run at line-rate to im plement ﬁrewalls, NAT, QoS, and to collect statistics. While each vendor’s ﬂow-table is diﬀerent, we’ve identiﬁed an interesting common set of functions that run in many switches and routers. OpenFlow exploits this common set of functions.OpenFlow provides an open protocol to program the ﬂow-table in diﬀerent switches and routers. A network administrator can partition traﬃc into production and research ﬂows. Researchers can control their own ﬂows -by choosing the routes their packets follow and the processing they receive. Inthis way, researchers can try new routing protocols, security models, addressing schemes, and even alternatives to IP. On the same network, the production traﬃc is isolated and processed in the same way as today.The datapath of an OpenFlow Switch consists of a Flow Table, and an action associated with each ﬂow entry. The set of actions supported by an OpenFlow Switch is extensible, but below we describe a minimum requirement for all switches. For high-performance and low-cost the data-path must have a carefully prescribed degree of ﬂexibility. This means forgoing the ability to specify arbitrary handling of each packet and seeking a more limited, but still useful, range of actions. Therefore, later in the paper, deﬁne a basic required set of actions for all OpenFlow switches.An OpenFlow Switch consists of at least three parts: (1) A Flow Table, with an action associated with each ﬂow entry, to tell the switch how to process the ﬂow, (2) A Secure Chan nel that connects the switch to a remote control process (called the controller), allowing commands and packets to be sent between a controller and the switch using (3) The OpenFlow Protocol, which provides an open and standard way for a controller to communicate with a switch. By specifying a standard interface (the OpenFlow Protocol) through which entries in the Flow Table can be deﬁned externally, the OpenFlow Switch avoids the need for researchers to program the switch.It is useful to categorize switches into dedicated OpenFlow switches that do not support normal Layer 2 and Layer 3 processing, and OpenFlow-enabled general purpose commercial Ethernet switches and routers, to which the Open-Flow Protocol and interfaces have been added as a new feature.Dedicated OpenFlow switches. A dedicated OpenFlow Switch is a dumb datapath element that forwards packets between ports, as deﬁned b y a remote control process. Figure 1 shows an example of an OpenFlow Switch.In this context, ﬂows are broadly deﬁned, and are limited only by the capabilities of the particular implementation of the Flow Table. For example, a ﬂow could be a TCP con nection, or all packets from a particular MAC address or IP address, or all packets with the same VLAN tag, or all packets from the same switch port. For experiments involving non-IPv4 packets, a ﬂow could be deﬁned as all packets matching a speciﬁc (but non-standard) header.Each ﬂow-entry has a simple action associated with it; the three basic ones (that all dedicated OpenFlow switches must support) are:1 Forward this ﬂow’s packets to a given port (or ports). This allows packets to be routed through the network. In most switches this is expected to take place at line-rate.2 Encapsulate and forw ard this ﬂow’s packets to a controller. Packet is delivered to Secure Channel, where it is encapsulated and sent to a controller. Typically used for the ﬁrst packet in a new ﬂow, so a controller can decide if the ﬂow should be added to the Flow Table. Or in some experiments, it could be used to forward all packets to a controller for processing.3 Drop this ﬂow’s packets. Can be used for security, to curb denial of service attacks, or to reduce spurious broadcast discovery traffic from end-hosts.An entry in the Flow-Table has three ﬁelds: (1) A packet header that deﬁnes the ﬂow, (2) The action, which deﬁnes how the packets should be processed, and (3) Statistics, which keep track of the number of packets and bytes foreach ﬂow, and the time since the last packet matched the ﬂow (to h elp with the removal of inactive ﬂows).In the ﬁrst generation “Type 0” switches, the ﬂow header is a 10-tuple shown in T able 1. A TCP ﬂow could be speciﬁed by all ten ﬁelds, whereas an IP ﬂow might not include the transport ports in its deﬁnition. Each header ﬁeld can be a wildcard to allow for aggregation of ﬂows, such as ﬂows in which only the VLAN ID is deﬁned would apply to all traﬃc on a particular VLAN.Table 1 The header ﬁelds matched in a “Type 0” OpenFlow switch.The detailed requirements of an OpenFlow Switch are defined by the OpenFlow Switch Specification [6].OpenFlow-enabled switches. Some commercial switches, routers and access points will be enhanced with the OpenFlow feature by adding the Flow Table, Secure Channel and OpenFlow Protocol (we list some examples in Section 5). Typically, the Flow Table will re-use existing hardware, such as a TCAM; the Secure Channel and Proto col will be ported to run on the switch’s operating system. Figure 2 shows a network of OpenFlow-enabled commercial switches and access points. In this example, all the Flow Tables are managed by the same controller; the OpenFlow Protocol allows a switch to be controlled by two or more controllers for increased performance or robustness. Our goal is to enable experiments to take place in an existing production network alongside regular traﬃc and applications. Therefore, to win theconﬁdence of network administrators, OpenFlow-enabled switches must isolate experimental traﬃc (processed by the Flow Table) from productiontraﬃc that is to be processed by the normal Layer 2 and Layer 3 pipeline of the switch. There are two ways to achieve this separation. One is to add a fourth action:4. Forward this ﬂow’s packets through the switch’s nor mal processing pipeline.The other is to deﬁne separat e sets of VLANs for experimental and production traﬃc. Both approaches allow normal production traﬃc that isn’t part of an experiment to be processed in the usual way by the switch. All OpenFlow-enabled switches are required to support one approach or the other; some will support both.Additional features. If a switch supports the header formats and the four basic actions mentioned above (and detailed in the OpenFlow SwitchSpeciﬁcation), then we call it a “Type 0” switch. We expect that many switches will support additional actions, for example to rewrite portions of the packet header (e.g., for NAT, or to obfuscate addresses on intermediate links), and to map packets to a priority class. Likewise, some Flow Tables will be able to match on arbitrary ﬁelds in the packet header, enabling experiments with new non-IP protocols. As a particular set of features emerges, we willdeﬁne a “Type 1” switch.Controllers.A controller adds and removes ﬂow-entries from the Flow Table on behalf of experiments. For example, a static controller might be a simple application running on a PC to statically establish ﬂows to interconnect a set of test computers for the duration of an experiment. In this case the ﬂows resemble VLANs in current networks— providing a simple mechanism toisolate experimental traﬃc from the production network. Viewed this way, OpenFlow is a generalization of VLANs.One can also imagine more sophisticated controllers that dynamicallyadd/remove ﬂows as an experiment progresses. In one usage model, a researcher might control the complete network of OpenFlow Switches and be free to decide how all ﬂows are processed.Figure 2 Example of a network of OpenFlow-enabled commercial switches and routers.A more sophisticated controller might support multiple researchers, each with diﬀerent accounts and permissions, enabling them to run multiple independent experiments on diﬀerent sets of ﬂows. Flows identiﬁed as under the control of a particular researcher (e.g., by a policy table running in a controller) could be delivered to a researcher’s user-level control program which then decides if a new ﬂow-entry should be added to the network of switches.3. USING OPENFLOWAs a simple example of how an OpenFlow Switch might be used imagine that Amy (a researcher) invented Amy-OSPF as a new routing protocol toreplace OSPF. She wants to try her protocol in a network of OpenFlow Switches, without changing any end-host software. Amy-OSPF will run in a controller; each time a new application ﬂow starts Amy-OSPF picks a route through a series of OpenFlow Switches, and adds a ﬂow-entry in each switch along the path. In her experiment, Amy decides to use Amy-OSPF for thetraﬃc entering the OpenFlow network from her own desktop PC— so she doesn’t disrupt the network for others. To do this, she deﬁnes one ﬂow to be all the traﬃc entering the Open-Flow switch through the switch port her PC is connected to, and adds a ﬂow-entry with the action “Enca psulate and forward all packets to a controller”. When her packets reach a controller, her new protocol chooses a route and adds a new ﬂow-entry (for the application ﬂow) to every switch along the chosen path. When subsequent packets arrive at a switch, they are processed quickly (and at line-rate) by the Flow Table. There are legitimate questions to ask about the performance, reliability and scalability of a controller that dynam ically adds and removes ﬂows as an experiment progresses: Can such a centralized controller be fast enough to process new ﬂows and program the Flow Switches? What happens when a controller fails? To some extent these questions were addressed in the context of the Ethane prototype, which used simple ﬂow switches and a central controller [7]. Preliminary results suggested that an Ethane controller based on a low-cost desktop PC could process over 10,000 new ﬂows per second —enough for a large college campus. Of course, the rate at which ne w ﬂows can be processed will depend on the complexity of the processing required by the re searcher’s experiment. But it gives us conﬁdence that mean ingful experiments can be run. Scalability and redundancy are possible by making acontroller (and the experiments) stateless, allowing simple load-balancing over multiple separate devices.3.1 Experiments in a Production NetworkChances are, Amy is testing her new protocol in a network used by lots of other people. We therefore want the network to have two additional properties:1 Packets belonging to users other than Amy should be routed using a standard and tested routing protocol running in the switch or router from a “name-brand” vendor.2 Amy should only be able to add ﬂow entries for her traﬃc, or for any traﬃc her network administrator has allowed her to control.Property 1 is achieved by OpenFlow-enabled switches. In Amy’s experiment, the default action for all packets that don’t come from Amy’s PC could be to forward them through the normal processing pipeline. Amy’s own packets would be forwarded directly to the outgoing port, without being processed by the normal pipeline.Property 2 depends on the controller. The controller should be seen as a platform that enables researchers to implement various experiments, and the restrictions of Property 2 can be achieved with the appropriate use of permissions or other ways to limit the powers of individual researchers to control ﬂow entries. The exact nature of these permission-like mechanisms will depend on how the controller is implemented. We expect that a variety of controllers will emerge. As an example of a concrete realization of a controller, some of the authors are working on a controller called NOX as a follow-on to the Ethane work [8]. A quite diﬀerent controller might emerge by extending the GENI management software to OpenFlow networks.3.2 More ExamplesAs with any experimental platform, the set of experiments will exceed those we can think of up-front — most experiments in OpenFlow networks are yet to be thought of. Here, for illustration, we oﬀer some examples of how OpenFlow-enabled networks could be used to experiment with new network applications and architectures.Example 1: Network Management and Access Con trol. We’ll use Ethane as our ﬁrst example [7] as it was the research that inspired OpenFlow. In fact, an OpenFlow Switch can be thought of as a generalization of Ethane’s datapath switch. Ethane used a speciﬁc implementation of a controller, suited for network management and control, that manages the admittance and routing of ﬂows. The basic idea of Ethane is to allow network managers to deﬁne a network-wide policy in the central controller, which is enforced directly by making admission control decisions for each new ﬂow. A controller checks a new ﬂow against a set of rules, such as “Guests can communicate using HTTP, but only via a web proxy” or “VoIP phones are not allowed to communicate with laptops.” A controller associates pack ets with their senders by managing all the bindings between names and addresses — it essentially takes over DNS, DHCP and authenticates all users when they join, keeping track of which switch port (or access point) they are connected to. One could envisage an extension to Ethane in which a policy dictates that particular ﬂows are sent to a user’s process in a controller, hence allowing researcher-speciﬁc processing to be performed in the network.Example 2: VLANs. OpenFlow can easily provide users with their own isolated network, just as VLANs do. The simplest approach is to statically declare a set of ﬂows which specify the ports accessible by traﬃc on a givenVLAN ID. Traﬃc identiﬁed as coming from a single user (for example, originating from speciﬁc switch ports or MAC addresses) is tagged by the switches (via an action) with the appropriate VLAN ID.A more dynamic approach might use a controller to manage authentication of users and use the knowledge of the users’ locations for tagging traﬃc at runtime.Example 3: Mobile wireless VOIP clients. For this example consider an experiment of a new call-handoﬀ mechanism for WiFi-enabled phones. In the experiment VOIP clients establish a new connection over the OpenFlow-enabled network. A controller is implemented to track the location of clients, re-routing connections — by reprogramming the Flow Tables — as users move through the network, allowing seamless handoﬀ from one access point to another.Example 4: A non-IP network. So far, our examples have assumed an IP network, but OpenFlow doesn’t require packets to be of any one format — so long as the Flow Table is able to match on the packet header. This would allow experiments using new naming, addressing and routing schemes. There are several ways an OpenFlow-enabled switch can support non-IP traﬃc. For example, ﬂows could be identiﬁed using their Ethernet header (MAC src and dst addresses), a new EtherType value, or at the IP level, by a new IP Version number. More generally, we hope that future switches will allow a controller to create a generic mask (oﬀset + value + mask), allowing packets to be processed in a researcher-speciﬁed way.Example 5: Processing packets rather than ﬂows.The examples above are for experiments involving ﬂows — where a controller makes decisions when the ﬂow starts. There are, of course, interesting experiments to be performed that require every packet to be processed. For example, an intrusion detection system that inspects every packet, an explicit congestion control mechanism, or when modifying the contents of packets, such as when converting packets from one protocol format to another.Figure 3: Example of processing packets through anexternal line-rate packet-processing device, such as a programmable NetFPGA router.There are two basic ways to process packets in an OpenFlow-enabled network. First, and simplest, is to force all of a ﬂow’s packets to pass through a controller. To do this, a controller doesn’t add a new ﬂow ent ry into the Flow Switch — it just allows the switch to default to forwarding every packet to a controller. This has the advantage of ﬂexibility, at the cost of performance. It might provide a useful way to test the functionality of a new protocol, but is unlikely to be of much interest for deployment in a large network.The second way to process packets is to route them to a programmable switch that does packet processing — for example, a NetFPGA-based programmable router. The advantage is that the packets can be processed at line-rate in a user-deﬁnable way; Figure 3 shows an example of how this could be done, in which the OpenFlow-enabled switch operates essentially as a patch-panel to allow the packets to reach the NetFPGA. In some cases, the NetFPGA board (a PCI board that plugs into a Linux PC) might be placed in the wiring closet alongside the OpenFlow-enabled switch, or (more likely) in a laboratory.4. THE OPENFLOW CONSORTIUMThe OpenFlow Consortium aims to popularize OpenFlow and maintain the OpenFl ow Switch Speciﬁcation. The Con sortium is a group of researchers and network administrators at universities and colleges who believe their research mission will be enhanced if OpenFlow-enabled switches are installed in their network.Membership is open and free for anyone at a school, college, university, or government agency worldwide. The OpenFlow Consortium welcomes individual members who are not employed by companies that manufacture or sell Ethernet switches, routers or wireless access points (because we want to keep the consortium free of vendor inﬂuence). To join, send email to .The Consortium web-site contains the OpenFlow Switch Speciﬁcation, a list of consortium members, and reference implementations of OpenFlow switches.Licensing Model: The OpenFlow Switch Speciﬁcation is free for all commercial and non-commercial use. (The exact wording is on the web-site.)Commercial switches and routers claiming to be “OpenFlow-enabled” must conform to the requirements of an OpenFlow Type 0 Switch, as deﬁned in the OpenFlow Switch Speciﬁcation. OpenFlow is a trademark of Stanford University, and will be protected on behalf of the Consortium.5. DEPLOYING OPENFLOW SWITCHESWe believe there is an interesting market opportunity for network equipment vendors to sell OpenFlow-enabled switches to the research community. Every building in thousands of colleges and universities contains wiring closets with Ethernet switches and routers, and with wireless access points spread across campus.We are actively working with several switch and router manufacturers who are adding the OpenFlow feature to their products by implementing a Flow Table in existing hardware; i.e. no hardware change is needed. The switches run the Secure Channel software on their existing processor.We have found network equipment vendors to be very open to the idea of adding the OpenFlow feature. Most vendors would like to support the research community without having to expose the internal workings of their products. We are deploying large OpenFlow networks in the Computer Science and Electrical Engineering departments at Stanford University. The networks in two buildings will be replaced by switches running OpenFlow. Eventually, all traﬃc will run over the OpenFlow network, with production traﬃc and experimental traﬃc being isolated on different VLANs under the control of network administrators. Researchers will control their own traﬃc, and be able to add/remove ﬂow-entries.。

光学专业词汇(2)

光学专业词汇大全retina视网膜Color Blind色盲weak color色弱Myopia-near-sighted近视Sensitivity to Light感光灵敏度boost推进lag behind落后于Hyperopic-far-sighted远视Dynamic Range动态范围critical fusionfrequency临界融合频率CFF临界闪变频率visual sensation视觉hromaticity Diagram色度图olor Temperature色温SV Model色彩模型hue色度aturation饱和度value纯度IE Model 相干红外能量模式omplementary Colors补色ar Pattern条状图形eat body 热稠化pproximate近似iolet紫罗兰ody Curve人体曲线olor Gamut色阶djacent邻近的ormal illumination法线照明rimary colors红黄蓝三原色olor saturation色饱和度olor Triangle颜色三角olor Notation颜色数标法Color Difference色差V Signal Processing电视信号处理amma Correction图像灰度校正onversion T ables换算表ut of balance失衡obble摇晃ack and forth前后lear(white)panel白光板ibrant震动uzzy失真uantum leap量子越迁VGA(800)600 derive from起源自ulprit犯人ender呈递nhibit抑制,约束tride大幅前进lemish污点bstruction障碍物cratch刮伤ubstance物质实质主旨esidue杂质riteria标准arameter参数djacent邻近的接近的synchrony异步luster串群utually互助得lgorithm运算法则hromatic Aberrations色差ovea小凹isual Acuity视觉灵敏度ontrast Sensitivity对比灵敏度emporal(time)Response反应时间endition表演,翻译nimation活泼又生气host重影arallax视差eficient缺乏的不足的isplay panel显示板G. Narrow Gauge)窄轨距ichroic mirror二色性的双色性的rewster Angle布鲁斯特角olarized Light极化光nternal reflection内反射irefringence 双折射xtinction Ratio 消光系数isalignment 未对准uarter Waveplates四分之一波片lemish污点瑕疵eometric几何学的ipple波纹apacitor电容器arallel平行的他antalum钽金属元素exsiccate使干燥xsiccate油管,软膏urnace炉子炉lectrolytic电解的,由电解产生的odule模数nalog类似物ut of the way不恰当incushion针垫拉ateral侧面得ectangle长方形ixture固定设备ontrol kit工具箱VI connector DVI数局线ertical垂直的horizontal 水平的nterlace隔行扫描ullion竖框直楞awtooth锯齿oggle套索钉eypad数字按键键盘angential切线iagnostic tool诊断工具agittal direction径向的ursor position光标位置ray aberration光线相差eighting factor权种因子ariables变量or now暂时,目前.眼下heck box复选框iry disk艾里斑xit pupil出[射光]瞳ptical path difference光称差ith respect to关于iffraction limited衍射极限avefront aberration波阵面相差pherical aberration球面象差araxial focus傍轴焦点hromatic aberration象差ocal coordinate system局部坐标系统oordinate system坐标系rthogonal直角得,正交的onic sections圆锥截面ccount for解决,得分arabolic reflector拋物面反射镜adius of curvature曲率半径pherical mirror球面镜eometrical aberration几何相差ncident radiation入射辐射lobal coordinate总体坐标n terms of根据按照eflected beam反射束YI= or your information供参考onstructive interference相长干涉hase difference相差chromatic singlet消色差透镜nterferometer干涉仪oundary constraint边界约束,池壁效应adii半径oom lenses变焦透镜eam splitters分束器iscrete不连续的,分离的bjective/ye lens物镜/目镜ainframe主机udimentary根本的,未发展的hotographic照相得摄影得axing繁重的,费力得lgebra代数学rigonometry三角学eometry几何学alculus微积分学hilosophy哲学agrange invariant拉格朗日不变量pherical球的ield information场信息tandard Lens标准透镜efracting Surface折射面stigmatism散光DTV高清晰度电视LV( Digital Light Valve)数码光路真空管，简称数字光阀iffraction grating衍射光珊ield angle张角araxial ray trace equations近轴光线轨迹方称ack focal length后焦距rincipal plane主平面ertex顶点,最高点stigmatism散光,因偏差而造成的曲解或错判edial中间的,平均的ariance不一致onic圆锥的,二次曲线ield of view视野ollimator瞄准仪onvolution回旋.盘旋,卷积uzzy失真,模糊berrated异常的symmetry不对称得ndicative可表示得arabolic拋物线得uffice足够,使满足pecification规格,说明书traightforward易懂的,直接了当的,olidify凝固,巩固.Constraints 约束,限制etrology度量衡ield coverage视场,视野ictate口述, 口授, 使听写, 指令, 指示, 命令, 规定rradiance发光, 光辉,辐照度erial空气得,空中得alide卤化物的onochromatic单色的,单频的olychromatic多色的spherical非球面的pherical球面的lignment列队,结盟ower(透镜放大率quiconvergence 同等收敛FL(effective focal length)有效焦距orkhorse广为应用的设备iconvex两面凸的lobal optimization整体最优化oncave凹得,凹面得ylindrical圆柱得olid model实体模型odulation Transfer Function调制传递函数n the heat of在最激烈的时候rotocol协议,规定riplet三重态anity心智健全inc锌,涂锌的elenide 硒化物,硒醚iscellaneous各色各样混在一起, 混杂的, 多才多艺的ersus与...相对olynomial多项式的oefficient系数xplicit function显函数distinct清楚的,截然不同的manate散发, 发出, 发源udimentary根本的,未发展的ntersection角差点RTE= araxial ray trace equation旁轴光线轨迹方程achromats 消色差透镜ardinal points基本方位eparations分色片ashed虚线low up放大verlay覆盖,覆盖图multiplayer 多层的umidity 湿度loat glass浮法玻璃quare one 出发点,端点quare up to 准备开打,坚决地面对eflecting telescope 反射式望远镜diagnostic tools诊断工具ayout plots规划图odulation transfer function调制转换功能FT快速傅里叶变换oint spread function点传播功能avelength波长ngle角度bsorption吸收ystem aperture系统孔径ens units透镜单位avelength range波长范围inglet lens单业透镜pectrum光谱iffraction grating衍射光栅sphere半球的DE= ens data editor Surface radius of curvature表面曲率半径urface thickness表面厚度aterial type材料种类emi-diameter半径ocal length焦距perture type孔径类型perture value孔径值ield of view视场icrons微米, d, and C=blue hydrogen,yellow helium,red hydrogen lines,primary wavelength主波长equential mode连续模式bject surface物表面he front surface of the lens透镜的前表面top光阑he back surface of the lens透镜的后表面he image surface像表面ymmetric相对称的iconvex两面凸的he curvature is positive if the center of curvature of the surface is to the right of the vertex.It is negative if the center of curvature is to the left of the vertex.如果曲率中心在最高点的右边,曲率值为正,如果曲率中心在最高点的左边,则曲率为负mage plane像平面ay Aberration光线相差angential direction切线方向agittal direction径向araxial focus旁轴的arginal边缘的pherical aberration球面像差ptimization Setup最优化调整ariable变量athematical sense数学角度FE= Merit Function Editor,Adding constraints增加约束ocal length焦矩长度perand操作数he effective focal length有效焦矩rimary wavelength主波长nitiate开始pot diagram位图表iry disk艾里斑xial chromatic aberration轴向色差ith respect to关于至于xit pupil出射光瞳PD= ptical path difference光学路径差iffraction limited衍射极限hromatic aberration色差hromatic focal shift色焦距变换araxial focus傍轴焦点xial spherical aberration轴向球差longitudinal spherical aberration 纵向球差:沿光轴方向度量的球差lateral spherical aberration垂轴球差在过近轴光线像点的垂轴平面内度量的球差coma、omatic aberration彗差eridional coma子午彗差agittal coma弧矢彗差stigmatism像散ocal coordinate system本地坐标系统eridional curvature of field子午场曲agittal curvature of field弧矢场曲ecentered lens偏轴透镜orthogonal直角的垂直的onic section圆锥截面ccount for说明,占有,得分tigmatic optical system无散光的光学系统ewtonian telescope牛顿望远镜arabolic reflector抛物面镜oci焦距hromatic aberration,色差uperpose重迭arabola抛物线pherical mirror球面镜MS=oot Mean Square均方根avefront波阵面pot size光点直径aussian quadrature高斯积分ectangular array矩阵列rid size磨粒度PSF= point Spread Function点扩散函数FT= fast Fourier Transform Algorithm快速傅里叶变换ross Section横截面bscurations昏暗ocal coordinates局部坐标系统ignette把…印为虚光照rrow key键盘上的箭头键efractive折射eflective反射n phase同相的协调的ray tracing光线追迹iffraction principles衍射原理rder effect式样提出的顺序效果nergy distribution能量分配onstructive interference相长干涉ispersive色散的inary optics二元光学hase advance相位提前chromatic single消色差单透镜iffractive parameter衍射参数oom lenses变焦透镜thermalized lenses绝热透镜nterferometers干涉计eam splitter分束器witchable component systems可开关组件系统ommon application通用ymmetry对称oundary constraint边界约束ulti-configuration(MC)MC Editor(MCE)perturbation动乱,动摇ndex accuracy折射率准确性ndex homogeneity折射率同种性ndex distribution折射率分配bbe number离差数esidual剩余的stablishing tolerances建立容差igure of merit质量因子olerance criteria公差标准odulation Transfer Function(MTF)调制传递函数oresight视轴，瞄准线Monte Carlo蒙特卡洛olerance operands误差操作数onic constant圆锥常数stigmatic aberration像散误差echanical tilt机械倾斜，机械倾角olerance Data Editor(TDE)公差资料编辑器ompensator补偿棱镜stimated system performance预估了的系统性能teratively反复的，重迭的tatistical dependence统计相关性equential ray trace model连续光线追迹模型mbed埋葬，埋入ultiple多样的，多重的，若干的on-Sequential Components不连续的组件orner cube角隅棱镜，三面直角透镜ensitivity Analysis灵敏度分析aceted reflector有小面的反射镜mit发射，发出est嵌套verlap交迭uter lens外透镜rute force强力eidel像差系数spect ratio长宽比RA边缘光线角RH边缘光线高度synchronous不同时的,异步Apodization factor变迹因子exapolar六角形ithered高频脉冲衍射调制传递函数（MTF），衍射实部传递函数（RTF），衍射虚部传递函数（ITF），衍射相位传递函数（PTF），方波传递函数（SWM）ogarithmic对数的arity奇偶longitudinal aberrations 纵向像差赛得系数:球差（PHA，I），彗差（OMA，2），像散（STI，3），场曲（CUR，4），畸变（IST，5），轴向色差（LA，L）和横向色差（TR，T）.横向像差系数:横向球差（SPH），横向弧矢彗差（SCO），横向子午彗差（TCO），横向弧矢场曲（SFC），横向子午场曲（TFC），横向畸变（DIS）横向轴上色差（LAC）。

科技词汇-光学专业英语

retina视网膜Color Blind色盲weak color色弱Myopia-near-sighted近视Sensitivity to Light感光灵敏度boost推进lag behind落后于Hyperopic-far-sighted远视Dynamic Range动态范围critical fusionfrequency临界融合频率CFF临界闪变频率visual sensation视觉hromaticity Diagram色度图olor Temperature色温SV Model色彩模型hue色度aturation饱和度value纯度IE Model 相干红外能量模式omplementary Colors补色ar Pattern条状图形eat body 热稠化pproximate近似iolet紫罗兰ody Curve人体曲线olor Gamut色阶djacent邻近的ormal illumination法线照明rimary colors红黄蓝三原色olor saturation色饱和度olor Triangle颜色三角olor Notation颜色数标法Color Difference色差V Signal Processing电视信号处理amma Correction图像灰度校正onversion Tables换算表ut of balance失衡obble摇晃ack and forth前后lear(white)panel白光板ibrant震动uzzy失真uantum leap量子越迁VGA(800)600 derive from起源自ulprit犯人ender呈递nhibit抑制,约束tride大幅前进lemish污点bstruction障碍物cratch刮伤ubstance物质实质主旨esidue杂质riteria标准arameter参数djacent邻近的接近的synchrony异步luster串群utually互助得lgorithm运算法则hromatic Aberrations色差ovea小凹isual Acuity视觉灵敏度ontrast Sensitivity对比灵敏度emporal(time)Response反应时间endition表演,翻译nimation活泼又生气host重影arallax视差eficient缺乏的不足的isplay panel显示板G. Narrow Gauge)窄轨距ichroic mirror二色性的双色性的rewster Angle布鲁斯特角olarized Light极化光nternal reflection内反射irefringence 双折射xtinction Ratio 消光系数isalignment 未对准uarter Waveplates四分之一波片lemish污点瑕疵eometric几何学的ipple波纹apacitor电容器arallel平行的他antalum钽金属元素exsiccate使干燥xsiccate油管,软膏urnace炉子炉lectrolytic电解的,由电解产生的odule模数nalog类似物ut of the way不恰当incushion针垫拉ateral侧面得ectangle长方形ixture固定设备ontrol kit工具箱VI connector DVI数局线ertical垂直的horizontal 水平的nterlace隔行扫描ullion竖框直楞awtooth锯齿oggle套索钉eypad数字按键键盘angential切线iagnostic tool诊断工具agittal direction径向的ursor position光标位置rayaberration光线相差eighting factor权种因子ariables变量or now暂时,目前.眼下heck box复选框iry disk艾里斑xit pupil出[射光]瞳ptical path difference光称差ith respect to关于iffraction limited衍射极限avefront aberration波阵面相差pherical aberration球面象差araxial focus傍轴焦点hromatic aberration象差ocal coordinate system局部坐标系统oordinate system坐标系rthogonal直角得,正交的onic sections圆锥截面ccount for解决,得分arabolic reflector物面反射镜adius of curvature曲率半径pherical mirror球面镜eometrical aberration几何相差ncident radiation入射辐射lobal coordinate总体坐标n terms of根据按照eflected beam反射束YI= or your information供参考onstructive interference相长干涉hase difference相差chromatic singlet消色差透镜nterferometer干涉仪oundary constraint边界约束,池壁效应adii半径oom lenses变焦透镜eam splitters分束器iscrete不连续的,分离的bjective/ye lens物镜/目镜ainframe主机udimentary根本的,未发展的hotographic照相得摄影得axing繁重的,费力得lgebra代数学rigonometry三角学eometry几何学alculus微积分学hilosophy哲学agrange invariant拉格朗日不变量pherical球的ield information场信息tandard Lens标准透镜efracting Surface折射面stigmatism散光DTV高清晰度电视LV( Digital Light Valve)数码光路真空管，简称数字光阀iffraction grating衍射光珊ield angle张角araxial ray trace equations近轴光线轨迹方称ack focal length后焦距rincipal plane主平面ertex顶点,最高点stigmatism散光,因偏差而造成的曲解或错判edial中间的,平均的ariance不一致onic圆锥的,二次曲线ield of view视野ollimator瞄准仪onvolution回旋.盘旋,卷积uzzy失真,模糊berrated异常的symmetry不对称得ndicative可表示得arabolic物线得uffice足够,使满足pecification规格,说明书traightforward易懂的,直接了当的,olidify凝固,巩固.Constraints 约束,限制etrology度量衡ield coverage视场,视野ictate口述, 口授, 使听写,指令, 指示, 命令, 规定rradiance发光, 光辉,辐照度erial空气得,空中得alide卤化物的onochromatic单色的,单频的olychromatic多色的spherical非球面的pherical球面的lignment列队,结盟ower(透镜放大率quiconvergence 同等收敛FL(effective focal length)有效焦距orkhorse广为应用的设备iconvex两面凸的lobal optimization整体最优化oncave凹得,凹面得ylindrical圆柱得olid model实体模型odulation Transfer Function 调制传递函数n the heat of在最激烈的时候rotocol协议,规定riplet三重态anity心智健全inc锌,涂锌的elenide 硒化物,硒醚iscellaneous各色各样混在一起, 混杂的, 多才多艺的ersus与...相对olynomial多项式的oefficient系数xplicit function显函数distinct清楚的,截然不同的manate散发, 发出, 发源udimentary根本的,未发展的ntersection角差点RTE= araxial ray trace equation旁轴光线轨迹方程achromats 消色差透镜ardinal points基本方位eparations分色片ashed虚线low up放大verlay覆盖,覆盖图multiplayer 多层的umidity 湿度loat glass浮法玻璃quare one 出发点,端点quare up to 准备开打,坚决地面对eflecting telescope 反射式望远镜diagnostic tools诊断工具ayout plots规划图odulation transfer function调制转换功能FT快速傅里叶变换oint spread function点传播功能avelength波长ngle角度bsorption吸收ystem aperture系统孔径ens units透镜单位avelength range波长范围inglet lens单业透镜pectrum光谱iffraction grating衍射光栅sphere半球的DE= ens data editor Surfaceradius of curvature表面曲率半径urface thickness表面厚度aterial type材料种类emi-diameter半径ocal length焦距perture type孔径类型perture value孔径值ield of view视场icrons微米, d, and C=blue hydrogen,yellow helium,red hydrogen lines,primary wavelength主波长equential mode连续模式bject surface物表面he front surface of the lens透镜的前表面top光阑he back surface of the lens透镜的后表面he image surface像表面ymmetric相对称的iconvex两面凸的he curvature is positive if thecenter of curvature of thesurface is to the right of the vertex.It is negative if the center of curvature is to the left of the vertex.如果曲率中心在最高点的右边,曲率值为正,如果曲率中心在最高点的左边,则曲率为负mage plane像平面ay Aberration光线相差angential direction切线方向agittal direction径向araxial focus旁轴的arginal边缘的pherical aberration球面像差ptimization Setup最优化调整ariable变量athematical sense数学角度FE= Merit Function Editor, Adding constraints增加约束ocal length焦矩长度perand操作数he effective focal length有效焦矩rimary wavelength主波长nitiate开始pot diagram位图表iry disk艾里斑xial chromatic aberration轴向色差ith respect to关于至于xit pupil出射光瞳PD= ptical path difference光学路径差iffraction limited衍射极限hromatic aberration色差hromatic focal shift色焦距变换araxial focus傍轴焦点xial spherical aberration轴向球差longitudinal sphericalaberration 纵向球差:沿光轴方向度量的球差lateral spherical aberration垂轴球差在过近轴光线像点的垂轴平面内度量的球差coma、omatic aberration彗差eridional coma子午彗差agittal coma弧矢彗差stigmatism像散ocal coordinate system本地坐标系统eridional curvature of field子午场曲agittal curvature of field弧矢场曲ecentered lens偏轴透镜orthogonal直角的垂直的onic section圆锥截面ccount for说明,占有,得分tigmatic optical system无散光的光学系统ewtonian telescope牛顿望远镜arabolic reflector抛物面镜oci焦距hromatic aberration,色差uperpose重迭arabola抛物线pherical mirror球面镜MS=oot Mean Square均方根avefront波阵面pot size光点直径aussian quadrature高斯积分ectangular array矩阵列rid size磨粒度PSF= point Spread Function 点扩散函数FT= fast Fourier Transform Algorithm快速傅里叶变换ross Section横截面bscurations昏暗ocal coordinates局部坐标系统ignette把…印为虚光照rrow key键盘上的箭头键efractive折射eflective反射n phase同相的协调的ray tracing光线追迹iffraction principles衍射原理rder effect式样提出的顺序效果nergy distribution能量分配onstructive interference相长干涉ispersive色散的inary optics二元光学hase advance相位提前chromatic single消色差单透镜iffractive parameter衍射参数oom lenses变焦透镜thermalized lenses绝热透镜nterferometers干涉计eam splitter分束器witchable componentsystems可开关组件系统ommon application通用ymmetry对称oundary constraint边界约束ulti-configuration(MC)MCEditor(MCE)perturbation动乱,动摇ndex accuracy折射率准确性ndex homogeneity折射率同种性ndex distribution折射率分配bbe number离差数esidual剩余的stablishing tolerances建立容差igure of merit质量因子olerance criteria公差标准odulation TransferFunction(MTF)调制传递函数oresight视轴，瞄准线Monte Carlo蒙特卡洛olerance operands误差操作数onic constant圆锥常数stigmatic aberration像散误差echanical tilt机械倾斜，机械倾角olerance Data Editor(TDE)公差资料编辑器ompensator补偿棱镜stimated systemperformance预估了的系统性能teratively反复的，重迭的tatistical dependence统计相关性equential ray trace model连续光线追迹模型mbed埋葬，埋入ultiple多样的，多重的，若干的on-Sequential Components不连续的组件orner cube角隅棱镜，三面直角透镜ensitivity Analysis灵敏度分析aceted reflector有小面的反射镜mit发射，发出est嵌套verlap交迭uter lens外透镜rute force强力eidel像差系数spect ratio长宽比RA边缘光线角RH边缘光线高度synchronous不同时的,异步Apodization factor变迹因子exapolar六角形ithered高频脉冲衍射调制传递函数（MTF），衍射实部传递函数（RTF），衍射虚部传递函数（ITF），衍射相位传递函数（PTF），方波传递函数（SWM）ogarithmic对数的arity奇偶longitudinal aberrations 纵向像差赛得系数:球差（PHA，I），彗差（OMA，2），像散（STI，3），场曲（CUR，4），畸变（IST，5），轴向色差（LA，L）和横向色差（TR，T）.横向像差系数:横向球差（SPH），横向弧矢彗差（SCO），横向子午彗差（TCO），横向弧矢场曲（SFC），横向子午场曲（TFC），横向畸变（DIS）横向轴上色差（LAC）。

电子专业学术语英文缩写简称对照表

CC CCD CCD CCD CCF CCFL CCFL(CCFT) CCTV CCTV CD CD CDCA CDDI CDES CDMA CDMA CDMA CDR CDVCC CF CFM CIF CIS CISPR CLNP CLP CM CM CM CMI CMISE CMOS CMRS CMTS COB
算术逻辑单元模拟用户线单元调幅管理模块隔位标志翻转接入网美国国家标准学会美国国家标准协会全光网络指信号仅在进出网络时才进行电/光和光/电的变 All Optical Network 换，而在网络中传输和交换的过程中始终以光的形式存在。 Automatic Protection Switching 自动保护倒换 Access and Remote Control 接入和遥控 Automati Slope Control 自动斜率控制 American standard code for information interchange 美国信息交换标准码 Application-Specific Integrated Circuits 专用集成电路 Advanced Technology Attachment 高级技术附加装置 pulse code 脉冲码 Analogue Trunk Unit 模拟中继单元 Asynchronous Transfer Mode 异步传输模式异步转移模式。将话音、图像、数据、视频等多种业务数字 Asynchronous Transfer Mode 化后转换成长度相同的分组（信元），包括信息域和元头，根据元头的信息进行传送。 Asynchrous Transfer Mode 异步传送方式 Administration Unit 管理单元 AU Pointer Positive Justification 管理单元正指针调整 Administrative Unit Alarm Indication SignalAU 告警指示信号 Administration Unit Group 管理单元组 Loss of Administrative Unit Pointer AU指针丢失 AU Pointer Negative Justification 管理单元负指针调整 Administration Unit Pointer 管理单元指针 audio visual 声视，视听 Auchio &Video Control Device 音像控制装置 American Wire Gauge 美国线缆规格 Bridge Amplifier 桥接放大器 Building Automation & Control net 建筑物自动化和控制网络

Moxa 网络监控移动能力启用白皮书说明书

WHITE PAPEREnabling Mobility in NetworkMonitoringYiwei ChenMoxa Product ManagerIntroductionEngineers face different challenges during each stage of the industrial network management lifecycle. During the installation stage, manual device configuration and testing is time consuming and prone to human error. During the operation stage, engineers are required to monitor network status in real time and minimize system downtime. During the maintenance stage, engineers often face long labor hours doing firmware upgrades or configuration changes on multiple devices. During the diagnostics stage, being able to quickly identify where critical network issues occur is essential. To help minimize the total cost of ownership, engineers are always on the lookout for new industrial network management tools that can help them overcome all of these challenges.Industrial network management software is usually installed in the control room, or is sometimes integrated with an existing SCADA system. But when you’re out of the co ntrol room or on the move, you could miss important messages such as network changes or errors, and fail to respond quickly enough. With the number of devices connected to industrial networks continually increasing, the ability to monitor and maintain your network—anytime, anywhere—is becoming more crucial than ever before to ensure that your operation is reliable and runs smoothly.Current statistics show that globally, the number of mobile users is now greater than the number of desktop users, and we can expect this global trend to expand into the industrial automation workplace. In fact, since engineers joining the workforce today are accustomed to using mobile devices in their private life, it is only natural that they would want to use the same devices to simplify their work life.In this white paper, we discuss the challenges in industrial network management and show how a mobile monitoring tool can help keep you informed of network status, even when you’re on the move. In addition, we’ll share experiences we’ve had helping customers from the rail industry reduce system downtime by utilizing the right mobile tools to quickly respond to network changes.Released on October 7, 2015© 2015 Moxa Inc. All rights reserved.Moxa is a leading manufacturer of industrial networking, computing, and automation solutions. With over 25 years of industry experience, Moxa has connected more than 30 million devices worldwide and has a distribution and service network that reaches customers in more than 70 countries. Moxa delivers lasting business value by empowering industry with reliable networks and sincere service for automation systems. Information about Moxa’s solutions is available at . You may also contact Moxa by email at *************.How to contact MoxaTel: 1-714-528-6777Fax: 1-714-528-6778Major Challenges in Industrial Network ManagementManaging a network can be a complex and often extensive operation, especially for industrial networks, and being able to monitor and manage devices is essential to ensuring that the network is running smoothly. However, with evolving business operations, administrators are often on the move, making it difficult to stay informed of or quickly respond to status changes in the network.When doing regular maintenance or troubleshooting at a field site where many network devices are deployed, engineers often face the daunting task of identifying specific devices hiding among a multitude of identical devices. Even with proper labeling and hardware placement, it can still take time to obtain the status information of a specific device onsite. As a result, faulty devices cannot be swapped out quickly enough to ensure that your operation runs smoothly.With the development of mobile networking tools, engineers can now improve operational efficiency and maximize network availability.Why Mobile Network Monitoring?Like their enterprise counterparts, automation engineers can now access their operational applications from mobile devices by installing an appropriate network monitoring app. The mobile network monitoring app is usually a client software tool designed to work in tandem with the network management software installed in the control room.The following diagram illustrates how a typical mobile app for network monitoring works to keep users informed of the ir network’s status. The app connects to the software server over an intranet or the Internet to access network status in real time. In addition, if the network is updated, the network management software server will send a push notification via the Apple cloud or Google cloud to alert the app user.A mobile phone app for network monitoring usually works as the client of the main network management software. Through the app, engineers can access the network status anytime,anywhere.How Mobile Networking Empowers Network OperatorsA mobile network monitoring app should support the following three features to ensure that monitoring a network from a mobile device is worth the effort.1.Sending Real-time AlertsWith a mobile network monitoring app, administrators can receive notifications of events pushed to their mobile devices. These real-time alerts allow administrators to take action immediately in response to critical events, even when they are out of the control room. For example, once an alert is received, they can contact maintenance engineers to do onsite troubleshooting and consequently reduce system downtime.2.Allowing Instant Network ChecksA mobile network app allows users to check the status of a network in real time. After youlog in to the app, it will inform you whether or not the network is operating normally. The app will also display detailed information of a specific network device, keeping network administrators in the know while they are on the move or out of the control room.Information, such as a device’s IP address, MAC address, location, and firmware version can be viewed from the app. For example, if an engineer receives an alert for a link-down event, they can readily access the information needed to determine which port is faulty.3.Finding Field Devices QuicklyIn certain scenarios, it could take a long time to manually search for a specific device from racks and racks of similar devices. Moreover, if automation engineers need to access the parameters or settings of a specific device for onsite troubleshooting, they would need to physically connect the device to a laptop computer using a web console or CLI (command line interface), or physically read the MAC address or serial number printed on the device, and then check the information with the computer. Either way, the engineer could end up spending much more time than would be necessary if the same information could bechecked using a mobile device.To make the task easier and more efficient, mobile network monitoring apps now usually come with a function that allows users to quickly find a particular device, and even view detailed device information.For example, each network device could be encoded with a unique QR code based on its MAC address. If the mobile phone app supports a built-in QR code scanner, engineers can scan the device’s QR code onsite to pull up information about that device, without needing to boot up a laptop computer or entering a device ID manually.With Moxa’s MXview ToGo app, users can not only scan the device to get detailed information, they can also activate the Device Locator function to find the device—which works by causing the device’s LED to blink in a way that is easy to recognize.Success StoryDeploying a Server/Client Solution for Industrial Network MonitoringTo ensure that a network operates reliably, industrial network management software is usually installed in large-scale networks in mission-critical industries, such as transportation, mining, and oil & gas. In this section, we share a success story from a railway application that uses a fiber Ethernet backbone built for data transmission between several stations located across a wide area. Since the application involves multiple control rooms spread over a wide area, the industrial network management software and the mobile phone app can help engineers access network status in real time and then respond quickly, thereby greatly reducing system downtime.This high-speed railway operator built a fiber Ethernet backbone for data transmission between its Operation Management Center and other railway stations to ensure high network availability. The customer used about 30 Moxa industrial rackmount switches (IKS-G6524) to connect to the pre-existing Layer 3 networks, and used the MXstudio industrial network management suite across the network management lifecycle, including for installation, operation, maintenance, and diagnostics. The MXstudio suite includes the MXview industrial network management software, MXconfig industrial network configuration tool, and N-Snap network snapshot tool.The railway operator’s network administrators recounted that they sometimes needed to leave the control room for patrol inspections within and around the station. Since MXview was already installed in the control room, they could install Moxa’s MXview ToGo mobile app, which works as a client of MXview, and then easily check the latest network status from their mobile phones. The dashboard design of the app makes it easy for engineers to tell whether the network is operating under Normal, Warning, or Critical conditions. In one notable incident, an IT engineer received a push notification about a downed link, used the app to determine wherethe broken link was located, and also connected to the MXview server to determine the cause. After determining the cause, the engineer contacted onsite staff immediately, allowing them to get the network link back up and running in no time.The diagram shows that engineers on the move can still get real-time network status with themobile app.ConclusionThe use of effective network management applications can help network administrators accomplish tasks efficiently during different stages of the network management lifecycle. With the changing business environment and improvements in mobile device technology, a mobile app for network monitoring allows administrators to be efficient, effective, and responsive when monitoring and maintaining an industrial network.Using a mobile app for network monitoring, administrators can view device and network status and receive real-time alerts from their mobile devices while on the move. In the field, administrators can quickly search for any device and view that device’s detailed configuration parameters with the click of a button.∙Learn more about Moxa’s MXview ToGo mobile app here:/MXview_ToGo∙Scan the following QR code to download the MXview ToGo app:iPhone OS Android OSDisclaimerThis document is provided for information purposes only, and the contents hereof are subject to change without notice. This document is not warranted to be error-free, nor subject to any other warranties or conditions, whether expressed orally or implied by law, including implied warranties and conditions of merchantability, or fitness for a particular purpose. We specifically disclaim any liability with respect to this document and no contractual obligations are formed either directly or indirectly by this document.。

关于传播学的国外文献

关于传播学的国外文献以下是关于传播学的一些国外文献的例子：1. McQuail, D. (2010). Mass communication theory. Sage Publications.（《大众传播理论》）2. Hall, S. (1980). Encoding/decoding. In S. Hall, D. Hobson, A. Lowe, & P. Willis (Eds.), Culture, media, language: Working papers in cultural studies, 1972-79 (pp. 128-138). Hutchinson.（《编码/解码》）3. Gitlin, T. (1980). The whole world is watching: Mass media in the making and unmaking of the New Left. University of California Press.（《全世界都在看：大众媒体对新左派的塑造和解体》）4. McCombs, M., & Shaw, D. L. (1972). The agenda-setting function of mass media. Public Opinion Quarterly, 36(2), 176-187.（《大众媒体的议程设置功能》）5. Jenkins, H. (2006). Convergence culture: Where old and new media collide. New York University Press.（《融合文化：新旧媒体的碰撞》）6. Couldry, N. (2012). Media, society, world: Social theory and digital media practice. Polity.（《媒体、社会、世界：社会理论与数字媒体实践》）7. Habermas, J. (1989). The structural transformation of the public sphere: An inquiry into a category of bourgeois society. MIT Press.（《公共领域的结构转型：对资产阶级社会范畴的探究》）8. Meyrowitz, J. (1985). No sense of place: The impact of electronic media on social behavior. Oxford University Press.（《无处感：电子媒体对社会行为的影响》）这些文献代表了传播学领域的经典和重要研究，涵盖了传播理论、媒体影响、社会媒体等方面的内容。

Kindle Paperwhite电子书阅读器显示屏的光学解构说明书

Optical teardown of a Kindle Paperwhite displayby OCTBart Johnson, Walid Atia, Mark Kuznetsov, Noble Larson, Eric McKenzie, Vaibhav Mathur,Brian Goldberg, Peter WhitneyAxsun Technologies, 1 Fortune Drive, Billerica, MA 01821, USAAbstract:An optical teardown, or reverse engineering, of an AmazonKindle Paperwhite electrophoretic display was performed by OpticalCoherence Tomography at 1060 nm. The display incorporates an opticaldiffuser, lightguide and scattering layers for white light illumination,capacitive touch sensing, and an electrophoretic display. All these layerscan be imaged by OCT as well as the thin film transistor array on the backside for driving the pixels. Phase sensitive OCT is used to measure motionof the pigment particles as the display changes between black and white.1. IntroductionThe Amazon Kindle Paperwhite [1] is an advanced electronic book reader that features a black and white electrophoretic display [2] as well as capacitive multitouch screen capability and an internal lightguide with an optical scattering layer [3,4] to provide uniform, white light illumination. Teardowns, where a new electronic gadget is disassembled and photographed are commonly found on technically oriented web sites. Here we perform an “optical teardown” of the first generation Kindle Paperwhite display by Optical Coherence Tomography (OCT). This highlights an industrial application of OCT in the 1060 nm wavelength range using advanced swept laser sources and data acquisition hardware [5].The heart of the Paperwhite is the electrophoretic display manufactured by E Ink Corporation [2]. Microencapsulated electrophoretic displays were first developed at MIT [6] and then refined [7] and commercialized [2]. The electrophoretic display layers consist of microcapsules of liquid containing black and white particles. The white particles are permanently charged negative and the black particles positive. They can thus be moved through electrophoresis by an electric field. Moving the white particles to the top of the capsule and black to the bottom results in a white pixel. Reversing the electric field turns it black. An interesting feature of these displays is that with no external electric field applied they remain in the same state. An image can be retained almost indefinitely with no power applied to the display. This fact, and the fact that the display emits no light, only reflects light, means that electrical power consumption is very low.Figure 1. Diagram of a microcapsule electrophoretic display.2. Overall display constructionThe Kindle Paperwhite display is a complex layered device of glass, plastic and thin films, in addition to the electrophoretic pigment microcapsules. A cross-section of the display taken with a 1060 nm OCT system is shown in Figure 2 and three dimensional renderings are shown in Figure 3. The top layer is a light diffuser to remove reflected light glare. It is followed by a high index light guide layer that conducts light from four white LEDs across the display. A graded surface of microprinted scattering centers [3,4] directs white light out of the guide layer to uniformly illuminate the pigment layer. Underneath the light guide is an 18 x 14 thin film capacitive touch screen layer. The capacitive array pitch is 7.0 x 6.6 mm on a 122 x 90 mm display area. The pigment layer contains fluid-filled microcapsules with charged black and white pigment particles. The capsules are sandwiched between a transparent ground plane and thin film transistor (TFT) driven pixel electrodes unseen below the pigment layer.Figure 2. Cross section image (left) and single A-line (right) of a Kindle Paperwhite display by OCT at 1060nm.Figure 3. Three dimensional renderings of OCT data showing the (a) diffuser,(b) scattering layer, (c) touch-screen capacitance layer, (d) E Ink pigmentlayer.3. Light diffuserFrom close examination of Figure 2, the light diffuser appears to be four thermoplastic layers containing small air bubbles. Microscope photos in Figure 4 show one image focused on the diffuser surface and another focused on the pigment layer below. The first image shows the diffuser bubbles and the second shows individual pixels in the pigment layer, slightly blurred because of the diffuser. The pixel pitch is 120 microns. We speculate that the diffuser material is polymethylmethacrylate (PMMA), which can be formulated with varying levels ofbubble content.on the E Ink pigment layer (right).4. Lightguide illumination layersThe internal white light illumination is a new feature in the Paperwhite model. Four white LEDs shine into a planar waveguide and microprinted scattering centers are designed to uniformly scatter light out of the guide onto the pigment layer. The success of this scheme is shown in the illumination uniformity map of Figure 5. The white/black contrast of the electrophoretic display is 17:1. These false color images were obtained by photographing the display using a RAW file format and removing the gamma compression from the final bitmap images to register linear light intensity.Figure 5. False color light intensity maps for a black page with whiteborder (left) and white page (right). Four white LEDs inject light into thelightgide at the bottom of the display and it propagates towards the top,being scattered onto the electrophoretic display along the way.The scattering centers are apparently imprinted on the surface of the high-index guide layer as described in these patents [3,4]. Each scattering center is a small island corrugated plastic as shown in Figure 6. The islands are smaller at the bottom of the display, near the white LEDs. For uniform illumination, the scattering needs to increase near the top of the display as the white light is depleted. These scattering islands are seen in the OCT image in Figure 3b as a kind of row-like structure in the coarser pitch dimension. These scattering centers are visible on a disassembled Paperwhite near the edge of the display where the centers overlay the metal film on the capacitance layer. The pitch of the centers is roughly 200 x 100 microns, although there appears to be some randomness of placement and ofcorrugation angle. Figure 7 shows the increased island size at the top of the display, away from the four illumination LEDs at the bottom.We speculate that the lightguide material is polycarbonate (PC), which is capable of being microprinted, as is done in the case of pressed compact discs, for example. OCT measurements of the optical and physical thickness of these layers indicates the refractive index is greater than 1.5, consistent with published numbers for PC at a 1 micron wavelength. This high index would provide light guiding when sandwiched between lower index materials.Figure 6. Microscope photograph of light emitted from the edge of thelightguide of a disassembled Paperwhite with the four illumination LEDsout of focus in the background (left). Structure of the lightguide andscattering centers (right).Figure 7. Microscope photographs of the microprinted scattering centers.The image is not clear because of the intervening diffuser layer. Closeexamination shows the corrugations from the microprinting.5. Touch screen capacitor layerThe capacitance layer shown in Figure 3c appears to be corrugated like the lightguide, but that is an artifact of the OCT imaging. The display was tilted in the OCT experiment to eliminate strong specular reflections. The result is that the specular beam from the capacitance layer misses the detector, but will get to the detector by scattering in the layer above as drawn in Figure 8. This modulates scattering islands onto the capacitance layer signal giving the mistaken impression that the capacitance layer is corrugated as well.The reflections from small metal traces circled in Figure 3c are real. Those are apparently metal crossover traces connecting one dimension of the touch screen projected capacitance grid. They are imperceptible to a person reading an E-book, but can be seenunder a microscope. We were unable to image the capacitor traces, likely made from indium tin oxide, with standard OCT and even phase sensitive OCT. The Paperwhite has a multi-touch capability and we are presuming that the capacitance layout is similar to that drawn in Figure 9.Figure 8. Optical path from the capacitance layer to the OCT detector,showing how the scattering layer affects the signal and confuses theinterpretation of the images.Figure 9. Diamond-shaped transparent pads in a typical projectedcapacitance array used for multi-touch sensing. The crossover tracesimaged by our OCT system are shown in black. The dark blue and dark redmetal traces that connect to the presumed ITO pads can be seen under amicroscope and by OCT.6. E Ink electrophoretic display layersOCT can image characters formed on the E Ink electrophoretic display layer, although the contrast at 1060 nm is not high. This is shown in Figure 10 in three dimensions along with reflections of all the intervening layers of the display.Figure 10. Three dimensional rendering of OCT data (left) and photograph(right) of the Kindle Paperwhite display with text from [8] displayed.While the display state, white/black/gray, is static with the power removed, the pigment particles exhibit Brownian motion. The particles are held in a semi-permanent potential well, but they are still free to move somewhat within the liquid. This can be seen from the speckle pattern in the M-mode image of Figure 11 that shows 250 A-lines from a stationary beam on the Kindle display. All the display layers and interfaces are seen, with unchanging speckle patterns except for the E Ink pigment layer. The light diffuser, for example, shows stable horizontal speckle lines over the 5 minute measurement period. The E Ink pigment speckle pattern changes for each A-line, showing that the pigment particles are in motion. This particular measurement was taken from a Kindle Paperwhite display in the white state whose battery was removed some months before.Figure 11. M-mode image formed from 250 A-lines taken over a 5 minutetime period. Stable speckle patterns are seen everywhere except withinthe E Ink pigment layer, showing that the pigment particles, while trappedin a stable potential well, still exhibit Brownian motion.An electrophoretic display has fluid-filled microcapsules containing mobile, charged pigment particles that can be driven towards or away from the viewer by electrophoretic forces depending on the direction of an electric field across the capsule layer. The Kindle Paperwhite has black and white particles. The motion of the white particles can be tracked by phase sensitive OCT since they are highly reflecting. Movement towards or away from the display surface can be detected, but lateral motion cannot since it produces no Doppler shift. Therefore motion due to electrophoresis can be measured, but motion from dielectrophoretic forces [9] cannot.The Kindle Paperwhite was mounted under a stationary beam from a 1060 nm, 100 kHz swept source [5]. A 12-bit data acquisition board [5] was used to stream 8.2 seconds of data to computer memory over a PCIe interface. While the data converters are capable of 550 MS/s rates, the board was actually clocked from a k-clock interferometer at frequencies ranging from 170 to 330 MHz. The fiber-based Mach-Zehnder k-clock drifts with temperature and the starting wavelength of the laser sweep jitters a few clock pulses sweep to sweep. These are not good conditions for a phase sensitive measurement and 2π phase errors can easily be made. We use a new phase unwrapping algorithm [10] that is tolerantto phase jitter to combat that problem. There are both one and two dimensional versions of the method.Figure 12 shows the results of an experiment where the Kindle Paperwhite display started white, was switched to black, and then back to white. Doing that required swiping the touch-screen display with a finger to “turn the page.” Physically touching the display moved it by about a dozen microns, even though the unit was firmly strapped down. Phase data from both the display surface and the pigment layer were collected so the overall device motion could be subtracted to just obtain the pigment motion. This creates a “virtual” common path interferometer, which is more pha se stable. This worked out fairly well, as shown in Figure 12. The red curve shows about 4 microns of white particle motion away from the display surface when it is switched to the black state. This is in the face of about 12 microns of overall motion from a finger pushing the display away from the OCT probe. A few 2π phase errors, which amount to about 0.5 microns, are likely, but overall the measurement looks reasonable.The Kindle Paperwhite is sluggish, taking about 0.5 seconds to respond to a finger command. The display reflectivity changes in synchronism with the pigment movement, not the finger motion. The 1060 nm laser signal from the pigment layer suffers from speckle effects and is very “noisy”. A simultaneous measurement of white light re flectivity shows a much cleaner trace. However, both measurements track and are synchronized with the pigment motion.It is clear that the white particles move away from the surface when the display transitions to the dark state, and the data shows this. The registered motion is only 3 microns, although the E Ink microcapsules are believed to be much bigger. Quantifying the movement in microns is problematic since there will be some signal from the stationary microcapsule walls and from the black particles that move in the opposite direction. It is expected that the signal will be dominated by white particle reflection only when they are near the top surface of the microcapsules. A more sophisticated model of the optical interaction is needed to probe further.Figure 12. Phase sensitive displacement measurement of pigmentparticles (top) and display reflectivity (bottom).7. Thin film transistorsA thin-film transistor (TFT) array drives the pixels in the electrophoretic display. The TFTs are not accessible optically from the front of the display since the pigment layer prevents light from penetrating that far. The TFTs can be imaged from the back of the display, nondestructively, by disassembling the Paperwhite. The TFTs are fabricated on a glass substrate and the layers deposited on the substrate can be imaged, although most of the deposited materials are metal and light does not penetrate further.Figure 13 shows three images of a 3x3 cell portion of the TFT pixel driver array. The cell pitch is 120 microns. The OCT images are of limited use because of the poor lateral resolution, however there is potential for added diagnostic information with a higher resolution scanner, especially with phase sensitive imaging. The phase sensitive image was made by subtracting the TFT layer phase from the substrate phase and applying the two-dimensional version of the filtered phase unwrapping algorithm of Ref. [10].Figure 13. 3x3 cells of the thin film transistor array seen with phasesensitive OCT (left), standard OCT (center), and white light microscopy(right)8. SummaryThis work illustrates an industrial application of OCT and reveals the impressive electro-optical technology behind the Kindle Paperwhite E-book reader. These measurements were made nondestructively. The reader was taken apart, but still worked when reassembled. An advanced 1060 nm swept source and a new data acquisition board capable of streaming large data sets made this work possible. These experiments also demonstrate a new algorithm for phase-unwrapping [10] that makes phase sensitive measurements possible in the face of modest laser phase jitter.References and links1.“Light Reading: How the Kindle Paperwhite Works,” NY Times, December 26, 2012,/interactive/2012/12/26/technology/light-reading.html?ref=personaltech&_r=2&2. E Ink Corporation. /3.L. Hatjasalo, K. Rinko, “Light panel with improved diffraction,” US Patent 6,773,126, Aug. 10, 2004.4.K. Rindo, “Ultra thin lighting element,” US Patent 7,565,054, Jul. 21, 2009.5.1060 nm swept source and prototype data acquisition board designed and manufactured by AxsunTechnologies. /6. B. Comiskey, J. D. Albert, H. Yoshizawa and J. Jacobson, “An electrophoretic ink for all-printed reflectiveelectronic displays,” Nature 394, 253-255 (1998).7. A. Loxley and B. Comiskey, “Capsules for electrophoretric displays and methods for making the same,” USPatent 6,262,833, Jul. 17,2001.8.Graham Farmelo, “The Strangest Man: The Hidden Life of Paul Dirac, Quantum Genius”, Kindle edition, (E-book location 3102).9.K.R. Amundson, A.C. Arango, J.M. Jacobson, T.H. Whitesides, M.D. McCreary, R.J. Paolini, Jr., “Methods fordriving electrophoretic displays using dielectrophoretic forces,” US Patent 7,999,787, Aug. 16, 2011.10.M.A. Navarro, J.C. Estrada, M. Servin, J.A. Quiroga, J. Vargas, “Fast two-dimensional simultaneous phaseunwrapping and low-pass filtering,” Optics Express 20, 2556 (2012).。

麻省理工学院的研究所如何做科研

麻省理工学院的研究所如何做科研双语版中文翻译版：网上流传的版本很多,其中的一个(中文版)/WangBNU/archive/2006/06/08/51255.html麻省理工学院人工智能实验室AI Working Paper 316 1988年10月来自MIT人工智能实验室：如何做研究？作者：人工智能实验室全体研究生编辑：David Chapman版本：1.3时间：1988年9月译者：柳泉波北京师范大学信息学院2000级博士生摘要本文的主旨是解释如何做研究。

我们提供的这些建议，对做研究本身（阅读、写作和程序设计），理解研究过程以及开始热爱研究（方法论、选题、选导师和情感因素），都是极具价值的。

不像正式论文那样，会列出所有的参考文献。

1. 简介这是什么？并没有什么神丹妙药可以保证在研究中取得成功，本文只是列举了一些可能会有所帮助的非正式意见。

目标读者是谁？本文档主要是为MIT人工智能实验室新入学的研究生而写，但对于其他机构的人工智能研究者也很有价值。

即使不是人工智能领域的研究者，也可以从中发现对自己有价值的部分。

如何使用？要精读完本文，太长了一些，最好是采用浏览的方式。

很多人觉得下面的方法很有效：先快速通读一遍，然后选取其中与自己当前研究项目有关的部分仔细研究。

本文档被粗略地分为两部分。

第一部分涉及研究者所需具备的各种技能：阅读，写作和程序设计，等等。

第二部分讨论研究过程本身：研究究竟是怎么回事，如何做研究，如何选题和选导师，如何考虑研究中的情感因素。

很多读者反映，从长远看，第二部分比第一部分更有价值，也更让人感兴趣。

.. 小节2 如何通过阅读打好AI研究的基础。

列举了重要的AI期刊，并给出了一些阅读的诀窍。

.. 小节3 如何成为AI研究领域的一员：与相关人员保持联系，他们可以使你保持对研究前沿的跟踪，知道应该读什么材料。

第1章空间经济现象及其理论渊源

Nagoya Kyushu Shanghai
Guangzhou
Manila
Kuala Lumpur
ASEAN
Jakarta
5
Global production share in terms of units
Automobiles
East Asia 37.1%
Japan , China
( 16.4
8.6 )
China ASEAN
世界的工厂
final goods (parts)
US (and the ROW)
中国三大地带（东、中、西）
100000 90000 80000 70000 60000
50000 40000 30000 20000 10000
0
东部：52.34％中部：31.21％西部：16.45％
第一章空间经济现象和经济理论
第一节空间经济现象第二节空间经济学的地位第三节空间经济学理论渊源
第一节空间经济现象第二节空间经济学的地位第三节空间经济学理论渊源
世界经济的一个重要现象是“集聚”
空间的尺度： 1、城市 2、大都市区 3、城市群 4、国家内部区域 5、国家 6、由多个国家组成的区域组织：如欧盟、北美、东亚等 7、全世界
1、空间经济学被主流经济学忽视的事实和原因
空间经济学长期被忽视的事实：长期以来，空间（或区位）未进入经济学的研究视野（经济学对空间和区位是视而不见、不闻不问），这是一个不争的事实！
“现在主流经济学中几乎没有任何空间分析”；
“例如，看看最新的经济学原理教材：约瑟夫·斯蒂格里茨（Joseph Stiglitz）的《经济学》。这是一本受到广泛好评的教材，要是有点美中不足的话就是它无所不包，全书有1100多页。然而索引中根本没有出现‘区位’ 或‘空间经济学’这样的词，仅仅在一处提到‘城市’，那是在讨论欠发达国家中农村人口向城市迁移时出现的。”

麻省理工大学

建筑管理学物理
Achievements
世界上第一台能够实时处理资料的“ 时处理资料的“旋风电脑” 电脑” B-2幽灵隐形战略轰炸机世界上第一个有人类感情的机器人Kismet 感情的机器人Kismet 目前21世纪培养诺贝目前21世纪培养诺贝尔奖得主最多的大学尔奖得主最多的大学（过去六年共16位得（过去六年共16位得主）；高科技武器和美国最高科技武器和美国最高机密的林肯实验室高机密的林肯实验室 ……
MIT
Location History School life Famous people from MIT Position
Address and transportation
位于麻萨诸塞州的波士顿，查尔斯河位于麻萨诸塞州的波士顿，查尔斯河（Charles River）将其与波士顿的后湾区 River）将其与波士顿的后湾区（Back Bay）隔开。 Bay）隔开。
School life
麻省理工学院占地168英亩麻省理工学院占地168英亩中央校区由一组互相连通的大楼组成宽带无线网络遍布校园各个角落，共有宽带无线网络遍布校园各个角落，共有 3000个收讯点 3000个收讯点
Get an overview of the campus
MIT FAST Thinking on March 5 MIT’ MIT’s Festival of Art + Science + Technology (FAST) FAST Thinking, a daylong exploration of Thinking, the convergence of music, language, technology and brain science. MIT FAST Thinking is part of MIT FAST — a three-monththree-month-long festival celebrating MIT’s MIT’ 150th birthday and its revolutionary work at the intersection of art, science, a在2007，2008,2009，2011年万维网的世 2007，2008,2009，2011年万维网的界大学排名里，MIT以压倒性的胜利排名界大学排名里，MIT以压倒性的胜利排名世界第一美国国家研究协会把MIT在美国大学的知美国国家研究协会把MIT在美国大学的知名度排第一至2010年，MIT已连续十八年在美国大学 2010年，MIT已连续十八年在美国大学理工学院排名夺冠美国Princeton Review在2006把MIT命名为美国Princeton Review在2006把MIT命名为全美最难进的大学

戴尔易安信 PowerEdge 服务器深度学习性能扩展白皮书说明书

WhitepaperDeep Learning Performance Scale-outRevision: 1.2Issue Date: 3/16/2020AbstractIn this whitepaper we evaluated the training performance of Scale-out implementation with the latest software stack and compared it with the results obtained in our previous paper [0]. Using TensorFlow as the primary framework and tensorflow benchmark models, the performance was compared in terms of throughput images/sec on ImageNet dataset, at a single node and multi-node level. We tested some of the more popular neural networks architectures for this comparison and demonstrated the scalability capacity of Dell EMC PowerEdge Servers powered by NVIDIA V100 Tensor Core GPUs.RevisionsDate Description3/16/2020Initial release AcknowledgementsThis paper was produced by the following:NameVilmara Sanchez Dell EMC - Software EngineerBhavesh Patel Dell EMC - Distinguished EngineerJosh Anderson Dell EMC - System Engineer (contributor)We would like to acknowledge:❖Technical Support Team - Mellanox Technologies❖Uber Horovod GitHub Team❖Nvidia Support teamTable of ContentsMotivation (4)Test Methodology (4)PowerEdge C4140-M Details (6)Performance Results - Short Tests for Parameter Tuning (8)Performance Results - Long Tests Accuracy Convergence (16)Conclusion (17)Server Features (18)Citation (19)References (19)Appendix: Reproducibility (20)MotivationWith the recent advances in the field of Machine Learning and especially Deep Learning, it’s becoming more and more important to figure out the right set of tools that will meet some of the performance characteristics for these workloads. Since Deep Learning is compute intensive, the use of accelerators like GPU become the norm, but GPUs are premium components and often it comes down to what is the performance difference between a system with and without GPU. In that sense Dell EMC is constantly looking to support the business goals of customers by building highly scalable and reliable infrastructure for Machine Learning/Deep Learning workloads and exploring new solutions for large scale distributed training to optimize the return on investment (ROI) and Total Cost of Ownership (TCO).Test MethodologyWe have classified TF benchmark tests in two categories: short and long tests. During the development of the short tests, we experimented with several configurations to determine the one that yielded the highest throughput in terms of images/second, then we selected that configuration to run the long tests to reach certain accuracy targets.Short TestsThe tests consisted of 10 warmup steps and then another 100 steps which were averaged to get the actual throughput. The benchmarks were run with 1 NVIDIA GPU to establish a baseline number of images/sec and then increasing the number of GPUs to 4 and 8. These tests allow us to experiment with the parameter tuning of the models in distributed mode.Long TestsThe tests were run using 90 epochs as the standard for ResNet50. This criterion was used to determine the total training time on C4140-M servers in distributed mode with the best parameter tuning found in the short tests and using the maximum number of GPUs supported by the system. In the section below, we describe the setup used, and Table 1gives an overall view on the test configuration.Testing SetupCommercial Application Computer Vision - Image classificationBenchmarks code▪TensorFlow Benchmarks scriptsTopology ▪ Single Node and Multi Node over InfiniBandServer ▪ PowerEdge C4140-M (4xV100-16GB-SXM2)Frameworks▪ TensorFlow with Horovod library for Distributed Mode[1] Models ▪ Convolutional neural networks: Inception-v4, vgg19,vgg16, Inception-v3, ResNet-50 and GoogLeNetBatch size ▪ 128-256GPU’s▪ 1-8Performance Metrics▪ Throughput images/second▪ Training to convergence at 76.2% TOP-1 ACCURACY Dataset ▪ ILSVRC2012 - ImageNetEnvironment ▪ DockerSoftware StackThe Table shows the software stack configuration used to build the environment to run the tests shown in paper [0] and the current testsSoftware Stack Previous Tests Current TestsTest Date February 2019 January 2020OS Ubuntu 16.04.4 LTS Ubuntu 18.04.3 LTSKernel GNU/Linux 4.4.0-128-genericx86_64GNU/Linux 4.15.0-69-genericx86_64nvidia driver 396.26 440.33.01CUDA 9.1.85 10.0 cuDNN 7.1.3 7.6.5 NCCL 2.2.15 2.5.6 TensorFlow 1.10 1.14 Horovod 0.15.2 0.19.0 Python 2.7 2.7 Open MPI 3.0.1 4.0.0 Mellanox OFED 4.3-1 4.7-3 GPUDirect RDMA 1.0-7 1.0-8Single Node - Docker Container TensorFlow/tensorflow:nightly-gpu-py3nvidia/cuda:10.0-devel-ubuntu18.04Multi Node -Docker Container built from nvidia/cuda:9.1-devel-ubuntu16.04nvidia/cuda:10.0-devel-ubuntu18.04Benchmark scripts tf_cnn_benchmarks tf_cnn_benchmarksDistributed SetupThe tests were run in a docker environment. Error! Reference source not found. 1 below shows the different logical layers involved in the software stack configuration. Each server is connected to the InfiniBand switch; has installed on the Host the Mellanox OFED for Ubuntu, the Docker CE, and the GPUDirect RDMA API; and the container image that was built with Horovod and Mellanox OFED among other supporting libraries. To build the extended container image, we used the Horovod docker file and modified it by adding the installation for Mellanox OFED drivers [2]. It was built from nvidia/cuda:10.0-devel-ubuntu18.04Figure 1: Servers Logical Design. Source: Image adapted fromhttps:///docs/DOC-2971Error! Reference source not found.2 below shows how PowerEdge C4140-M is connected via InifniBand fabric for multi-node testing.Figure 2: Using Mellanox CX5 InfiniBand adapter to connect PowerEdge C4140 in multi-nodeconfigurationPowerEdge C4140-M DetailsThe Dell EMC PowerEdge C4140, an accelerator-optimized, high density 1U rack server, is used as the compute node unit in this solution. The PowerEdge C4140 can support four NVIDIA V100 Tensor Core GPUs, both the V100-SXM2 (with high speed NVIDIA NVLink interconnect) as well as the V100-PCIe models.Figure 3: PowerEdge C4140 ServerThe Dell EMC PowerEdge C4140 supports NVIDIA V100 with NVLink in topology ‘M’ with a high bandwidth host to GPU communication is one of the most advantageous topologies for deep learning. Most of the competitive systems, supporting either a 4-way, 8-way or 16-way NVIDIAVolta SXM, use PCIe bridges and this limits the total available bandwidth between CPU to GPU. See Table 2 with the Host-GPU Complex PCIe Bandwidth Summary.ConfigurationLink Interface b/n CPU-GPU complex Total BandwidthNotesM4x16 Gen3128GB/sEach GPU has individual x16 Gen3 to Host CPUTable 2: Host-GPU Complex PCIe Bandwidth SummarySXM1SXM2SXM4SXM3CPU2CPU1x16x16x16x16UPI Configuration MX16 IO SlotX16 IO SlotFigure 4: C4140 Configuration-MNVLinkPCIePerformance Results - Short Tests for Parameter Tuning Below are the results for the short tests using TF 1.14. In this section we tested all the models in multi node mode and compared the results obtained with TF 1.10 in 2019. Throughput CNN Models TF 1.10 vs TF 1.14The Figure 5 several CNN models comparing results with TF 1.10 vs TF 1.14. In Figure 6 we notice that the performance gain is about 1.08X (or 8%) between the two releases.Figure 5: Multi Node PowerEdge C4140-M – Several CNN Models TF 1.10 vs TF 1.14Figure 6 Multi Node PowerEdge C4140-M – Several CNN Models TF 1.10 vs TF 1.14 (Speedup factor)Performance Gain with XLASince there was not much performance gain with the basic configuration, we decided to explore the limits of GPU performance using other parameters. We looked at XLA (Accelerated Linear Algebra) [3], by adding the flag –xla=true at the script level. By default, the TensorFlow graph executor “executes” the operations with individual kernels, one kernel for the multiplication, one kernel for the addition and one for the reduction. With XLA, these operations are “fused” in just one kernel; keeping the intermediate and final results in the GPU, reducing memory operations, and therefore improving performance.See below Figure 7 for results and Figure 8 for speedup factors across the models. The models inception-v4, inception-v3 and ResNet-50 showed much better performance using XLA, with speedup factors from 1.35X up to 1.43X. Since the ResNet-50 model is most widely used, we used it to continue the rest of the tests.Figure 7: Multi Node PowerEdge C4140-M. Several CNN Models TF 1.10 vs TF 1.14 + XLAFigure 8: Multi Node PowerEdge C4140-M. Several CNN Models TF 1.10 vs TF 1.14 + XLA (Speedupfactor)ResNet-50’s Performance with TF 1.14 + XLAIn this section, we evaluated the performance of ResNet-50 model trained with TF 1.14 and TF 1.14 with XLA enabled. The tests were run with 1 GPU, 4 GPUs, and 8 GPUs and the results were compared with those obtained for version TF 1.10 from our previous paper [0]. Also, we explored the performance using batch size of 128 and 256. See Figure 9 and Figure 10.Figure 9: Multi Node PowerEdge C4140-M. ResNet-50 BS 128 TF 1.10 vs TF 1.14 vs TF 1.14 + XLAAs we saw in the previous section ResNet-50 with batch size 128 with 8 GPUs had a performance gain of ~3% with TF 1.10 vs TF 1.14, and ~35% of performance gain with TF 1.10 vs TF 1.14 with XLA enabled, see Figure 9. On the other hand, ResNet-50 with batch size 256 with 8 GPUs had a performance gain of ~2% with TF 1.10 vs TF 1.14, and ~46% of performance gain with TF 1.10 vs TF 1.14 with XLA enabled, see Figure 10. Due to the higher performance of ResNet-50 with batch size 256, we have selected it to further optimize performance.Figure 10: Multi Node PowerEdge C4140-M. ResNet-50 BS 256 TF 1.10 vs TF 1.14 vs TF 1.14 + XLAResNet-50 with TF 1.14 + XLA + GPUDirect RDMAAnother feature explored in our previous paper was GPUDirect RDMA which provides a direct P2P (Peer-to-Peer) data path between GPU memory using a Mellanox HCA device between the nodes. In this test, we enabled it by adding the NCCL flag – x NCCL_NET_GDR_LEVEL=3 at the script level (this variable replaced the variable NCCL_IB_CUDA_SUPPORT in NCCL v 2.4.0). NCCL_NET_GDR_LEVEL variable allows you to control when to use GPUDirect RDMA between a NIC and a GPU. Example level 3 indicates to use GPUDirect RDMA when GPU and NIC are on the same PCI root complex [4].Figure 11: Multi Node PowerEdge C4140-M. ResNet-50 with TF 1.14 + XLA + GPUDirect RDMA Figure 11shows the results of ResNet-50 with TF 1.14 w/XLA enabled, with and without GPUDirect RDMA. We did not observe much performance gains using GPUDirect RDMA across nodes i.e. the performance remained the same and hence we did not explore it further in our testing. This is not to say that GPUDirect RDMA does not help when using scale-out, all we are saying is we did not see the performance gains; hence we are not exploring it further in this paper.ResNet-50’s configuration for better performanceFigure 12 summarizes the results of different configurations explored for the short tests and based on the tests we found that the best performance was achieved using the combination below:ResNet-50 + BS 256 + TF 1.14 + XLA enabledFigure 12: Multi Node PowerEdge C4140-M - ResNet-50’s Configuration for Best PerformanceResNet-50’s Scale-outPowerEdge C4140 using Nvidia 4x NVLink architecture scales relatively well when using Uber Horovod distributed training library and Mellanox InfiniBand as the high-speed link between nodes. It scales ~3.9x times within the node and ~6.9x using scale-out for ResNet-50 with batch size 256. See Figure 13.Figure 13: Multi Node PowerEdge C4140-M - ResNet-50’s Scale-out vs Scale-upFigure 14: Multi Node PowerEdge C4140-M vs CompetitorThe above benchmarks shown in Figure 14 are done on 2 servers C4140 x4 V100 GPUs, each connected by Mellanox ConnectX-5 network adapter with 100Gbit/s over IPoIB. The Dell EMC distributed mode with Horovod achieved 85% of scaling efficiency for ResNet-50 batch size 256 compared with the ideal performance; on the other hand, it achieved 95% of scaling efficiency versus a test run by TF team on 2018 with a VM (virtual machine) instance on GCP(Google cloud) with 8x V100 GPUs and batch size=364 [5].Performance Results - Long Tests Accuracy Convergence Our final tests were to determine the total training time for accuracy convergence with the latest tensorflow version.In this section we decided to include all the batch sizes we tested in our previous paper and compared it with ResNet-50 using batch size 256.Figure 15shows the total training time achieved when running ResNet-50 with different batch sizes and both versions of tensorflow (TF 1.10 vs TF 1.14 with XLA enabled).On average using TF 1.14 +XLA was ~1.3X faster than our previous tests.Figure 15: Multi Node PowerEdge C4140-M - ResNet-50’s Long Training for Accuracy Conv.Conclusion•The performance with TF 1.14 among the models was just slightly superior (~1%-8%) versus TF 1.10. On the other hand, TF 1.14 with XLA boosted the performance up to~46% among the models ResNet-50, Inception-v3 and Inception-v4.•In the case of ResNet-50 model, its performance improved up to ~ 3% with TF 1.14, and up to ~46% with TF 1.14 and XLA enabled. ResNet-50 batch size 256 scaled better(1.46X) versus ResNet-50 BS 128 (1.35X).•The configuration with the highest throughput (img/sec) was ResNet-50 batch size 256 trained with distributed Horovod + TF 1.14 + XLA enabled.•Dell EMC PowerEdge C4140 using Nvidia 4x NVLink architecture scales relatively well when using Uber Horovod distributed training library and Mellanox InfiniBand as the high-speed link between nodes. It scale-out ~3.9X times within the node and ~6.9X across nodes for ResNet-50 BS 256.•On average, the training time for the long tests to reach accuracy convergence were ~ 1.3 X faster using distributed Horovod + TF 1.14 + XLA enabled.•There is a lot of performance improvement being added continuously either at the GPU level, library level or framework level. We are continuously looking at how we can improve our performance results by experimenting with different hyper parameters.•TensorFlow in multi-GPU/multi-node with Horovod Distributed and XLA support improve model performance and reduce the training time, allowing customers to do more with no additional hardware investment.Server FeaturesCitation@article {sergeev2018horovod,Author = {Alexander Sergeev and Mike Del Balso},Journal = {arXiv preprint arXiv: 1802.05799},Title = {Horovod: fast and easy distributed deep learning in {TensorFlow}}, Year = {2018}}References•[0] https:///manuals/all-products/esuprt_solutions_int/esuprt_solutions_int_solutions_resources/servers-solution-resources_white-papers52_en-us.pdf•[1] Horovod GitHub, “Horovod Distributed Deep Learning Training Framework” [Online].Available: https:///horovod/horovod•[2] Mellanox Community, “How to Create a Docker Container with RDMA Accelerated Applications Over 100Gb InfiniBand Network” [Online]. Available:https:///docs/DOC-2971•[3] TensorFlow, “XLA: Optimizing Compiler for Machine Learning” [Online] Available: https:///xla•[4] NCCL 2.5, “NCCL Environment Variables” [Online]. Available:https:///deeplearning/sdk/nccl-developer-guide/docs/env.html#nccl-ib-cuda-support•[5] TensorFlow, “Pushing the limits of GPU performance with XLA” [Online]. Available: https:///tensorflow/pushing-the-limits-of-gpu-performance-with-xla-53559db8e473Appendix: ReproducibilityThe section below walks through the setting requirements for the distributed Dell EMC system and execution of the benchmarks. Do this for both servers:•Update Kernel on Linux•Install Kernel Headers on Linux•Install Mellanox OFED at local host•Setup Password less SSH•Configure the IP over InfiniBand (IPoIB)•Install CUDA with NVIDIA driver•install CUDA Toolkit•Download and install GPUDirect RDMA at the localhost•Check GPUDirect kernel module is properly loaded•Install Docker CE and nvidia runtime•Build - Horovod in Docker with MLNX OFED support•Check the configuration status on each server (nvidia-smi topo -m && ifconfig && ibstat && ibv_devinfo -v && ofed_info -s)•Pull the benchmark directory into the localhost•Mount the NFS drive with the ImageNet dataRun the system as:On Secondary node (run this first):$ sudo docker run --gpus all -it --network=host -v /root/.ssh:/root/.ssh --cap-add=IPC_LOCK -v /home/dell/imagenet_tfrecords/:/data/ -v/home/dell/benchmarks/:/benchmarks -v /etc/localtime:/etc/localtime:ro --privileged horovod:latest-mlnxofed_gpudirect-tf1.14_cuda10.0 bash -c "/usr/sbin/sshd -p 50000; sleep infinity"On Primary node:$ sudo docker run --gpus all -it --network=host -v /root/.ssh:/root/.ssh --cap-add=IPC_LOCK -v /home/dell/imagenet_tfrecords/:/data/ -v/home/dell/benchmarks/:/benchmarks -v /etc/localtime:/etc/localtime:ro --privileged horovod:latest-mlnxofed_gpudirect-tf1.14_cuda10.0•Running the benchmark in single node mode with 4 GPUs:$ python /benchmarks/scripts/tf_cnn_benchmarks/tf_cnn_benchmarks.py --device=gpu --data_format=NCHW --optimizer=sgd --distortions=false --use_fp16=True --local_parameter_device=gpu --variable_update=replicated --all_reduce_spec=nccl --data_dir=/data/train --data_name=imagenet --model=ResNet-50 --batch_size=256 --num_gpus=4 --xla=true•Running the benchmark in multi node mode with 8 GPUs:$ mpirun -np 8 -H 192.168.11.1:4,192.168.11.2:4 --allow-run-as-root -xNCCL_NET_GDR_LEVEL=3 -x NCCL_DEBUG_SUBSYS=NET -x NCCL_IB_DISABLE=0 -mcabtl_tcp_if_include ib0 -x NCCL_SOCKET_IFNAME=ib0 -x NCCL_DEBUG=INFO -xHOROVOD_MPI_THREADS_DISABLE=1 --bind-to none --map-by slot --mca plm_rsh_args "-p 50000" python /benchmarks/scripts/tf_cnn_benchmarks/tf_cnn_benchmarks.py --device=gpu --data_format=NCHW --optimizer=sgd --distortions=false --use_fp16=True --local_parameter_device=gpu --variable_update=horovod --horovod_device=gpu --datasets_num_private_threads=4 --data_dir=/data/train --data_name=imagenet --display_every=10 --model=ResNet-50 --batch_size=256 --xla=True。

专业术语常用名词缩写中英文对照

BA:Bridge Amplifier 桥接放大器 TOP BAC:Building Automation & Control net 建筑物自动化和控制网络 BAM:Background Administration Module 后管理模块 BBER:Background Block Error Ratio 背景块误码比 BCC:B-channel Connect ControlB 通路连接控制 BD:Building Distributor BEF:Buiding Entrance Facilities 建筑物入口设施 BFOC:Bayonet Fibre Optic Connector 大口式光纤连接器 BGN:Background Noise 背景噪声 BGS: Background Sound 背景音响 BIP-N:Bit Interleaved Parity N code 比特间插奇偶校验 N 位码 B-ISDN:Brand band ISDN 宽带综合业务数字网 B-ISDN:Broad band -Integrated Services Digital Network 宽带综合业务数字网
专业术语常用名词缩写中英文对照请选择您所要查询的字母:A-B-C-D-E-F-G-H-I-L-M-O-P-R-S-T-U-V A:Actuator 执行器 A:Amplifier 放大器 A:Attendance 员工考勤 A:Attenuation 衰减 AA:Antenna amplifier 开线放大器 AA:Architectural Acoustics 建筑声学 AC:Analogue Controller 模拟控制器 ACD:Automatic Call Distribution 自动分配话务 ACS:Access Control System 出入控制系统 AD:Addressable Detector 地址探测器 ADM:Add/Drop Multiplexer 分插复用器 ADPCM:Adaptive Differential ulse Code Modulation 自适应差分脉冲编码调制 AF:Acoustic Feedback 声反馈 AFR:Amplitude /Frequency Response 幅频响应 AGC:Automati Gain Control 自动增益控制 AHU:Air Handling Unit 空气处理机组 A-I:Auto-iris 自动光圈 AIS:Alarm Indication Signal 告警指示信号 AITS:Acknowledged Information Transfer Service 确认操作 ALC:Automati Level Control 自动平衡控制 ALS:Alarm Seconds 告警秒 ALU:Analogue Lines Unit 模拟用户线单元 AM:Administration Module 管理模块 AN:Access Network 接入网 ANSI:American National Standards Institute 美国国家标准学会 APS:Automatic Protection Switching 自动保护倒换 ASC:Automati Slope Control 自动斜率控制 ATH:Analogue Trunk Unit 模拟中继单元 ATM：Asynchrous Transfer Mode 异步传送方式 AU- PPJE:AU Pointer Positive Justification 管理单元正指针调整 AU:Administration Unit 管理单元 AU-AIS:Administrative Unit Alarm Indication SignalAU 告警指示信号 AUG:Administration Unit Group 管理单元组

白皮书：PERC Point-to-Point Resistance（P2P resistance）

WHITE PAPER Introduction PERC Point-to-point resistance (P2P resistance) functionality is a crucial EDA technology to enable complex P2P effective resistance measurement along ESD paths in automation forfoundry qualified ESD/Latch-up checker or in-house custom checker. This technology is appliedto the entire chip, block, and IP designs on cell or transistor level layout database. Since theESD path count could grow to thousands or even ten thousand, it is vital that the ESD path-oriented R extraction and distributed matrix solving capability has outstanding performance.This technology does not estimate P2P resistance using shortest or longest path schemes,but instead uses an accurate simulation focused on the critical layout polygons of ESDpaths, including the P/G network. The P2P resistance measurement yields result in effectiveresistance (ohms) for each path measured. That result is a lumped resistance value reflecting allinterconnect polygon layers between Source (current injection) and Sink (current Sink). Althoughthe P2P resistance value gives users immediate information on how effective the ESD pathbehaves in discharging an ESD surge, a violated P2P resistance value alone is difficult for layoutengineer to act on for any layout fix. When these checks fail, the resistance is reported for thefailing source/sink pair, but a single resistance value does not guide how to fix it. A physical layerchange will need to occur to fix it, but that single value provides no guidance on where to lookand what to do, and that is especially problematic on complex paths that can span dozens ofphysical layers. Ultimately this can lead to design delays and even failing silicon. IC Validator PERCIC Validator™ PERC is part of the more prominent IC Validator physical verification solution. ICValidator provides industry-leading solutions for DRC, LVS, FILL, pattern matching, and manyother applications.IC Validator PERC leverages StarRC™ for R-extraction and Python for a rich programmingenvironment, and together those are the technology backbone for the flow. That flow can thenbe used for netlist checks, netlist driving layout DRC checks, P2P , and current density.AuthorsFrank FengDir, Business Development,Synopsys Jonathan White Dir, Applications Engineering, SynopsysDebugging Point-to-Point Resistance Using Contribution by Layer in IC Validator PERCIC ValidatorFigure 1: IC Validator Physical Verification SolutionIC Validator PERC is qualified for significant foundry ESD/LUP checking of P2P resistance measurement, even at the full-chip level. IC Validator PERC P2P flow employs StarRC for R extraction, which is the industry gold standard. To provide better P2P resistance analysis for the layout engineer to act when there is a P2P resistance violation, IC Validator PERC offers a distinctive featureto analyze P2P resistance result contribution by layer. This feature enables beneficial information for the user to decide which interconnect layers are high contributors to be the candidates for layout fix.Enabling IC Validator PERC Collecting Database for P2P Resistance Contribution by Layer To use this debug capability, R reduction will need to be turned off in the P2P resistance flow. The reason is that reduction greatly simplifies the R network, and information about the actual fractured layout polygons is lost. This control is in the StarRC tech file, and the IC Validator PERC flow has a mechanism to define rules to PERC that then get passed into StarRC. The “readme” for your foundry runset for P2P resistance would provide the information needed to do this.Performance Impact of Producing Data for P2P Resistance Contribution by LayerIf the user is only running PERC P2P on IO nets and other non-P/G nets, then enabling P2P resistance contribution by layer will have minimal impact on performance. However, if the user is running PERC P2P, including P/G nets, then the user should expecta performance impact, depending on the size of the design. This performance impact is insignificant if the design is small (chip size is less than a few mm^2). If the chip size is larger than 10 mm^2, the performance cost will be more significant due to StarRC not reducing the R network of the P/G nets. The chip size estimation for performance cost is a rough guide to keep in mind andnot a hard rule.Using IC Validator VUE to Access P2P Resistance Contribution by Layer in Conjunction with Layout HighlightingThe user can launch IC Validator PERC job as usual. Upon the successful completion of IC Validator PERC P2P job, the user will be able to analyze P2P resistance contribution by layer in IC Validator VUE together with a layout viewer supported by IC Validator VUE. In the below section, the debug process in conjunction with P2P resistance contribution by layer is described. The user starts a layout viewer such as IC Validator WorkBench, Virtuoso, or Custom Compiler™. With the layout database open, the user invokes IC Validator VUE and loads topcell.vue file. Select “PERC Errors” tab to point to IC Validator PERC P2P run results; the violation P2P paths associated with each check name are listed on the Violation Browser page (on the left panel of VUE main window). Select one of the paths, and more details are shown as Violation Details/Description (on the right panel of VUE main window). Select the top line for each path in the Violation Details panel, right-click to drop-down list of can-do function, select (left click) on “PERC Path Heatmap,” and then a “Highlight Path” dialog window pops out. In the top portion of “Highlight Path” dialog window, there are list of symbols. Select the rightmost symbol (looks like a table), the P2P resistance contribution by layer table named “Contributionto Path Resistance” is shown. Each of these steps in the VUE debug procedures to access P2P resistance contribution by layer is displayed in figure 2.Figure 2: A flow chart describes accessing P2P resistance contribution by layer in VUE.The P2P resistance contribution by layer data (named as Contribution to Path Resistance table) provides how the lumped total resistance of the selected P2P path is summed up from various layers. Since each layer has its sheet resistance, layout polygons alone can’t tell the user what to do. The top contributors of resistance combined with layer polygon highlight capability in the “Highlight Path” dialog window give the layout engineer a much better idea of what to do for a layout fix. Figure 3 shows one P2P path measured from ESD diodes one physical power Pad of a power net.©2021 Synopsys, Inc. All rights reserved. Synopsys is a trademark of Synopsys, Inc. in the United States and other countries. A list of Synopsys trademarks is availableat /copyright .html . All other names mentioned herein are trademarks or registered trademarks of their respective owners.Figure 3: Pictures show how P2P resistance contribution by layer table provides valuable information for a user to focus debug and layout fix priorityInterpreting Contribution by Layer Results to Fix Design IssuesFixing P2P resistance issues in a design can be a complex problem. Proper fixing depends most heavily on the designer’s knowledge of their design and what changes they can make to resolve it. Additional information like a contribution by layer is intended to help understand the results more effectively so that the designer can apply their design knowledge with more confidence and greater speed. So, for example, knowing that the top thick metal represents 70% of the P2P resistance contribution does show what to change with that metal routing. But it does indicate quickly to a design whether the results are as expected or whether something unusual has occurred. And it gives confidence for the designer to make changes to the top think metal knowing that there will not be unintended consequences for that changes.SummaryPoint-to-point resistance checking is an essential component of robust ESD design verification. However, debugging the reported errors can be a real challenge and frustration to ESD engineers. IC Validator PERC provides the “contribution by layer” feature in its P2P Heatmap interface to IC Validator VUE, which offers tremendous insight into fixing these critical design errors. This saves time in the design cycle and gives higher confidence going into silicon ESD testing.。

产业融合理论研究综述

经济研究产业融合是伴随技术变革与扩散过程而出现的一种新经济现象。

国外产业融合思想最早起源于美国学者罗森伯格（Ｒｏｓｅｎｂｅｒｇ１９６３）。

２０世纪７０年代末，该现象受到广泛关注，并最初来源于实业界关于“电脑和通信”融合图景的描绘。

此后扩展到学术界和政界。

１９７８年，麻省理工学院（ＭＩＴ）媒体实验室Ｎ・尼古路庞特对计算、印刷、和广播业三者间技术融合的模型化描述，并认为交叉处是增长最快，创新最多的地方［１］，开启了学术界对产业融合研究的大门。

此后的一段时间里，学术界对产业融合的研究成果也只是零星的出现。

直到２０世纪９０年代中后期，美国新电信法案通过后，信息通信领域里跨媒体、跨产业、跨地域的企业并购风起云涌之时，才出现了产业融合研究的高潮，大量研究文献开始出现。

纵观目前国内外对产业融合理论的研究，主要集中在对其内涵、类型、动力及意义几个方面。

一、产业融合的内涵国内外学者对产业融合现象进行研究时，尝试着从不同视角对其内涵进行界定，结果使得其内涵至今仍未形成一致意见。

具体而言，现有研究成果主要从技术、产品、企业、市场及制度等视角来界定产业融合内涵的。

（一）技术视角产业融合现象研究最早始于技术融合。

罗森伯格（Ｒｏｓｅｎｂｅｒｇ　１９６３）在对美国机器工具产业演化研究中发现同一技术向不同产业扩散的现象，并把该现象定义为“技术融合”［２］。

所谓技术融合是指迄今为止不同产业分享共同知识和技术基础的过程［３］（Ａｔｈｒｅｙｅ＆Ｋｅｅｂｌｅ　２０００；Ｆａｉ＆Ｔｕｎｚｅｌｍａｎｎ　２００１；Ｌｉｎｄ　２００４；张磊２００１）。

也是某些技术在一系列产业中的广泛应用和扩散，并导致创新活动发生的过程［４］（Ｓａｈａｌ　１９８５；Ｄｏｓｉ　１９８８；岭言　２００１；卢东斌２００１）。

当不同产业技术的一体化（即共享相同的技术基础）显著地影响或改变另一产业中产品、竞争、价值创造过程的本质（Ｌｅｉ２００１）时，意味着技术融合产生。

MIT牛人解说数学体系

MIT牛人解说数学体系在过去的一年中，我一直在数学的海洋中游荡，research进展不多，对于数学世界的阅历算是有了一些长进。

为什么要深入数学的世界作为计算机的学生，我没有任何企图要成为一个数学家。

我学习数学的目的，是要想爬上巨人的肩膀，希望站在更高的高度，能把我自己研究的东西看得更深广一些。

说起来，我在刚来这个学校的时候，并没有预料到我将会有一个深入数学的旅程。

我的导师最初希望我去做的题目，是对appearance和motion建立一个unified的model。

这个题目在当今Computer Vision中百花齐放的世界中并没有任何特别的地方。

事实上，使用各种Graphical Model把各种东西联合在一起framework，在近年的论文中并不少见。

我不否认现在广泛流行的Graphical Model是对复杂现象建模的有力工具，但是，我认为它不是panacea，并不能取代对于所研究的问题的深入的钻研。

如果统计学习包治百病，那么很多“下游”的学科也就没有存在的必要了。

事实上，开始的时候，我也是和Vision中很多人一样，想着去做一个Graphical Model——我的导师指出，这样的做法只是重复一些标准的流程，并没有很大的价值。

经过很长时间的反复，另外一个路径慢慢被确立下来——我们相信，一个图像是通过大量“原子”的某种空间分布构成的，原子群的运动形成了动态的可视过程。

微观意义下的单个原子运动，和宏观意义下的整体分布的变换存在着深刻的联系——这需要我们去发掘。

在深入探索这个题目的过程中，遇到了很多很多的问题，如何描述一个一般的运动过程，如何建立一个稳定并且广泛适用的原子表达，如何刻画微观运动和宏观分布变换的联系，还有很多。

在这个过程中，我发现了两个事情：我原有的数学基础已经远远不能适应我对这些问题的深入研究。

在数学中，有很多思想和工具，是非常适合解决这些问题的，只是没有被很多的应用科学的研究者重视。

2022年考研考博-考博英语-中国科学院考试全真模拟易错、难点剖析AB卷(带答案)试题号：22

2022年考研考博-考博英语-中国科学院考试全真模拟易错、难点剖析AB卷（带答案）一.综合题(共15题)1.单选题Toyota said on Tuesday that it will（） its annual year-end sales event early.问题1选项A.give awayB.fix onC.kick offD.bring up【答案】C【解析】考查动词短语辨析。

give away “放弃，分发”；fix on “确定，固定”；kick off “开始”；bring up“提出，养育”。

句意：星期二，丰田汽车公司宣布将提前开展其一年一度的年终促销活动。

选项C符合句意。

2.单选题I looked at Mum and thought（） she was as nice as she looked maybe all our lives would have been better.问题1选项A.only ifB.if onlyC.even ifD.even though【答案】B【解析】考察固定搭配。

only if只有；even if = even though即使；if only “要是……就”；句意为：我注视着母亲，心想：要是她像看上去那么和蔼可亲就好了，这样我们的生活也许会变得更好。

选项B符合句意。

3.单选题Mist steals silently in, turns familiar landscapes strange, dampens sounds, ________ vision—then clears suddenly and without warning.问题1选项A.spoilsB.blowsC.tearsD.blurs【答案】D【解析】【选项释义】A. spoils 破坏，糟蹋B. blows 吹，刮C. tears 撕裂，撕开D. blurs 视线模糊，看不清【考查点】动词辨析。

2022年考研考博-考博英语-中国科学院考试全真模拟易错、难点剖析AB卷(带答案)试题号：47

2022年考研考博-考博英语-中国科学院考试全真模拟易错、难点剖析AB卷（带答案）一.综合题(共15题)1.单选题The theory that business could operate totally without the aid of government has proved to be a （）.问题1选项A.allusionB.seclusionC.illusionD.confusion【答案】C【解析】考察名词辨析。

allusion “暗示”；seclusion “与世隔绝”；illusion “幻想；错觉”；confusion“混乱；困惑”。

句意：认为商业运营能够完全脱离政府援助的这种观点是一种幻想。

选项C 符合句意。

2.单选题The closer one can get to reality, the easier the learning by the student,（） that the necessary knowledge has been given previously to facilitate comprehension.问题1选项A.soB.despiteC.suchD.provided【答案】D【解析】考察连词辨析。

so 因此，despite 尽管，such 如此；provided that “如果，假设”;句意：如果先前已经学习了能够帮助理解的必要知识，那么距离现实越近，学习就会越容易。

选项D符合句意。

3.单选题The Egyptians used to believe that literacy was divine. Scholars no longer ________ that theory, but why ancient civilizations developed writing was a mystery for a long time.问题1选项A.embraceB.involveC.formulateprehend【答案】A【解析】【选项释义】A. embrace 接受，采纳B. involve 包含，牵涉C. formulate 制订，规划D. comprehend 理解，领悟【考查点】动词辨析。

关于非连文本的著作

关于非连文本的著作引言随着信息技术的发展和互联网的普及，传统的书籍已经不再是人们获取知识和信息的唯一途径。

与此同时，非连文本（Nonlinear Text）作为一种新兴的文本形式，逐渐受到人们的关注和喜爱。

本文将探讨非连文本的定义、特点以及其在著作中的运用，旨在帮助读者更好地理解和欣赏这种独特的文本形式。

什么是非连文本？非连文本是指在时间和空间上不按照传统线性顺序排列的文本形式。

与传统线性文本不同，非连文本具有多样化、多层次和多路径等特点，读者可以根据自己的兴趣和需求选择不同路径进行阅读。

非连文本可以包括但不限于图像、音频、视频、超链接等元素，并通过交互方式实现与读者之间的互动。

非连文本著作中的特点1.多媒体元素丰富：非连文本著作通常会使用多种媒体元素来丰富内容，如图片、音频、视频等。

这些元素可以直观地传达信息，提高读者的阅读体验和理解效果。

2.多层次结构：非连文本著作中的内容通常按照多层次结构进行组织。

读者可以通过点击或选择不同的链接路径来深入探索各个层次的内容，从而获得更全面、详细的信息。

3.交互性强：非连文本著作通过交互方式与读者互动，使阅读过程更加参与和有趣。

读者可以根据自己的兴趣和需求选择不同路径进行阅读，自由地探索和发现新知识。

4.非线性阅读体验：非连文本著作打破了传统线性阅读方式的限制，使得读者可以根据自己的兴趣和需求选择不同路径进行阅读。

这种非线性的阅读体验能够激发读者的思考和创造力，提供更广阔的思维空间。

非连文本著作在实践中的应用1.电子书籍：随着电子书籍的普及，非连文本著作成为一种新型书籍形式。

电子书籍通过图像、音频、视频等元素丰富了内容，并提供了交互式的阅读体验，使读者能够更加深入地理解和体验书籍内容。

2.教育教学：非连文本著作在教育教学中有着广泛的应用。

教师可以利用非连文本著作创作丰富多样的教材，通过多媒体元素和交互方式提升学生的学习兴趣和效果。

同时，学生也可以通过阅读非连文本著作来主动探索和发现知识，培养自主学习的能力。

Convergence

Concept of Convergence
17
Convergence of research traditions
– Receiver needs vs. sender needs
– Critical mass concept – tipping point
– A sociodynamic term to describe the existence of sufficient momentum in a social system such that the momentum becomes self-sustaining and fuels further growth
– URL everywhere; IP everything
Concept of Convergence
8
Milton Mueller
– Technological driver (cont.)
3. multimedia capability (mix of text, voice, video, graphics, animation)
Concept of Convergence
6
Don Tapscott
3. new compression and mass data storage methods
4. computer technology to manage such a complex nationwide or worldwide service
Computing
$$$$ Most Growth
Printing
Broadcasting
Concept of Converginesses of computing, (tele-)communications

1、下载文档前请自行甄别文档内容的完整性，平台不提供额外的编辑、内容补充、找答案等附加服务。
2、"仅部分预览"的文档,不可在线预览部分如存在完整性等问题,可反馈申请退款(可完整预览的文档不适用该条件!)。
3、如文档侵犯您的权益，请联系客服反馈,我们会尽快为您处理(人工客服工作时间：9:00-18:30)。

!
Convergence is a new paradigm that can yield critical advances in a broad array of sectors, from health care to energy, food, climate, and water.
!"#$%"#&$'()"#%*+'"($*,"-+$+.'/$#&0"#+1$ 2"(+*2+$+.&$345$6*/.'(7+"($8))'2&9$ :;<$= $>+?$@A1$>+&?$B=<$ 6*/.'(7+"(1$CD$;<<<;$ ;<;?E:F?=:;:$GG$B=E?H;I?FF:=$ $
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
M A S S A C H U S E T T S
I N S T I T U T E
O F
Байду номын сангаас
T E C H N O L O G Y
The Third Revolution:
The Convergence of the Life Sciences, Physical Sciences, and Engineering
/+
! "#$%&!'(!)'*+&*+,!
!"#$%&'(#!%")))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))*$ +),%-.$-!./,%0,(%"-.$1."(. )))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))2$ MN(NS.!JDFCNOD(G"!MNTF"U.DFSK <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<V$ !"#$%&'()$*#+,-.)&,/0$1,-#2.-3'$3/4$5#--.-3'$6&,-,78999999999999999999999999999999999999999999999999999999999999999999999999999:$ !"#$;#2,/4$*#+,-.)&,/0$<#/,=&2(9999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999>$ .@N!.@DMO!MNTF"U.DFS!8!(FSTNMWNS(N <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<X$ NYGCL"NK!FH!(FSTNMWNS(N!DS!JDFCNOD(G"!MNKNGM(@ <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<Z$ 5,=?.)3)&,/3-$6&,-,7