Intel® Xeon Phi™ processors x200 (formerly Knights Landing) are computational beasts. Their theoretical peak performance is up to 3 TFLOP/s and measured memory bandwidth is up to 490 GB/s. This performance is available without any difference in programming models compared to general-purpose x86-like CPUs.
Colfax is offering a free trial program for this technology. This program is available through Intel’s sponsorship. The Colfax Cluster has 64 compute nodes based on Intel Xeon Phi 7250 processors. Intel® Omni-Path fabric interconnects the nodes. This cluster is at your service for two weeks for testing and evaluation.
In this 1-hour webinar I will describe how you can get the most out of your two weeks on the cluster:
- What workloads you can run to see the performance
- How to prepare your own code to run on the cluster
- Where to learn the best optimization practices for this and similar architectures