Traditional InfiniBand* support involves MPI message transfer over the Reliable Connection (RC) protocol. While RC is long-standing and rich in functionality, it does have certain drawbacks: since it requires that each pair of processes setup a one-to-one connection at the start of the execution, memory consumption could (at the worst case) grow linearly as more MPI ranks are added and the number of pair connections grows.
We are approaching another wonderful hackathon in 3 weeks. This one is named the "Virtual Hackathon". It is virtual in the sense that we will not all be in 1 building, but in separate houses working on different things. We will communicate to the other houses through the interwebs when necessary. Some of us members [including me] will be offering our houses for 3 days, from May 24 to May 26.
This two day webinar series introduces you to the world of multicore and manycore computing with Intel® Xeon® processors and Intel® Xeon Phi™ coprocessors.
This page contains replays of 6 sessions covering a variety of topics as listed below:
When running an MPI program, process placement is critical to maximum performance. Many applications can be sufficiently controlled with a simple process placement scheme, while some will require a more complex approach. The Intel® MPI Library offers multiple options for controlling process placement within the Hydra process manager.
面向英特尔® MIC 架构进行应用的适用性分析
面向英特尔® MIC 架构的编译器方法
使用英特尔® MPI 实现并行化
Response to our February session of the Intel® Xeon® and Xeon® Phi™ Introduction to High Performance Application Development for Multicore and Manycore-Live webinar was gratifying and overwhelming, but finally we worked through all of your questions. Some of the questions required a context we lost with the transcript and some were only partially formed, or of special interest, or duplicates. We gathered together all the questions of general interest from the webinar and farmed them out to our experts for more complete answers. We'll assembled that list and sorted it by
This is currently a placeholder for Intel® Trace Analyzer and Collector usage guides. Until articles are added, please visit the Intel® Trace Analyzer and Collector product page. You can also view the documentation.
Filtering in the Intel® Trace Collector will apply specified filters to the trace collection process. This directly reduces the amount of data collected. The filter rules can be applied either via command line arguments or in a configuration file (specified by the environment variable VT_CONFIG). Filters are evaluated in the order they are listed.
Filters consist of three components, type, pattern, and body. The basic filter is of type STATE. The other types, SYMBOL and ACTIVITY, apply pattern replacements, but are otherwise identical to STATE.
