Tkool Electronics

The comprehensive line consists of enGage I, II and III, covering the full range of industrial vehicle instrumentation needs, from golf cars and lawn tractors to sophisticated material handling vehicles. The multi-functional gauges can be bulk ordered as blanks, then user customized with the desired capabilities, from tachometers and fuel gauges to temperature and field programmable maintenance monitoring. More than 10 functions can be selected in virtually any combination the OEM or reseller desires.

pickit 3

The comprehensive line consists of enGage I, II and III, covering the full range of industrial vehicle instrumentation needs, from golf cars and lawn tractors to sophisticated material handling vehicles. The multi-functional gauges can be bulk ordered as blanks, then user customized with the desired capabilities, from tachometers and fuel gauges to temperature and field programmable maintenance monitoring. More than 10 functions can be selected in virtually any combination the OEM or reseller desires.

Benchmark outline MPLS domains contain three distinct forwarding points: ingress, transit and egress (for more on MPLS, see MPLS raises bar for network speed and management”).

Each of these functional areas has its own unique processing tasks that can be implemented differently. Therefore, in order to get a complete picture of MPLS performance, the MPLS benchmark tests cover each area independently. Each has its own set of parameters, tests and reporting formats. This comprehensive approach minimizes the ability of participants to gear systems to one particular aspect of the MPLS specification.

pickit 3

The MPLS benchmark specifications are based on the existing benchmark methodology found in RFC2544 and the NPF Internet Protocol version 4 (IPv4) Forwarding Level Benchmark Implementation Agreement. The tests have been tailored to address the unique aspects of testing MPLS technology on network processor systems.

The MPLS IA is a performance-testing benchmark and not a verification of MPLS protocol conformance. Only the basic aspects of MPLS are tested in this implementation agreement to determine the relative performance of MPLS in network-processing systems. The MPLS benchmark specification does not exhaustively test every possible feature or parameter. In situations where multiple options exist that do not have significantly different performance impacts, the most widely implemented method was chosen as the performance metric for that group of functions.

For example, FEC implementations could be network prefix classifier, IP host address classifier or 5-6 tuple classifier. The longest-prefix-match (LPM) or network prefix classifier was selected for the benchmark because it is the most popular classifier in MPLS today.

pickit 3

Another important attribute of the MPLS benchmark spec is that it addresses only data-plane functionality. MPLS control-plane updates do not stress the system enough to affect data-plane forwarding performance. Thus the result of a control-plane benchmark test is not that useful to a vendor, and tests to measure control-plane performance were not included in the benchmark. Instead, an optional test to measure the maximum number of label switch paths (LSPs) that can be supported by an MPLS router at the throughput rate was added. This measure is useful to vendors evaluating an MPLS router for deployment in traffic engineering and flow management network solutions.

Test specifics The first step in testing a network-processing subsystem using the MPLS benchmark implementation agreement is to create a reference design identifying a specific device under test, or DUT (Figure 1) . This test design may consist of one or more media interfaces and a fabric interface. It may include multiple network processors and any number of co-processors connected to the network processor in any way. The choice of speed and media type is left to network processor vendors or customers comparing different network processors.

pickit 3

The reference design must detail the significant components of the DUT including a block diagram, component list and other elements such as mechanical size, media and fabric interfaces, ports, processor and coprocessor details, and memory.

At the same time, Agere is fielding four devices aimed at 3G wireless basestations and wireline access applications. Two of those, sold under the TAAD moniker, integrate embedded memory, an MCU, framers and ATM switching capabilities, functions that usually take up to 12 chips, the company said. The 1,600-mm2 devices consume 3 watts, less than half the power it takes using separate devices.

Using the same architecture, Agere has also spun out two devices-the SAR-1K and SAR-500-that perform segmentation-and-reassembly chores in voice gateway and DSLAM equipment. Available now, the TAAD and SAR chips range from $125 to $300 each in 10,000-unit lots.

The TAAD and SAR devices show how Agere is trying to get the most out of its hardware and software development projects. Many portions of the software tools are the same, while some parts of the hardware were deactivated for certain devices to cut test costs and reduce the final price.

With one major design, we were able to come out with four devices and four cost points,” said product-marketing manager Jaime Mitchell.

http://www.eet.com

访客,请您发表评论:

Powered By Tkool Electronics

Copyright Your WebSite.sitemap