Xeon Scalable Processors
Chris Myers, Steve Lantz
Cornell Center for Advanced Computing
Revisions: 4/2024, 3/2023, 1/2022, 2/2021 (original)
Intel Xeon Scalable Processors, such as the first-generation Skylake and second-generation Cascade Lake chips, come in a variety of different models, with different technical specifications. Both sets of processors are designed for a wide range of HPC workloads, support Intel AVX-512 instructions, and are equipped with two vector processing units (VPUs) for every core. The technical improvements introduced by Intel in the first two generations will be discussed in some detail, with a special focus on the models deployed by TACC in the Stampede3 and Frontera supercomputers.
Objectives
After you complete this topic, you should be able to:
- Identify the key features of Intel Xeon Scalable Processors
- Distinguish between first-generation Skylake and second-generation Cascade Lake processors
- Explain Intel's motivation for giving Cascade Lake the ability to do lower-precision vector arithmetic for neural network algorithms
- Discuss the importance of multiple layers of parallelism in applications
Prerequisites
- Familiarity with High Performance Computing (HPC) concepts. Those who are less conversant with HPC terms and techniques should be prepared to inspect the glossary terms rather frequently. It may also be helpful to review Cornell Virtual Workshop content on Parallel Programming Concepts and High-Performance Computing and either MPI or OpenMP.
- Readers who need an introduction to either Stampede3 or Frontera will find it helpful to first review one of more of the following items: the Stampede3 User Guide, the Frontera User Guide, and the Getting Started on Frontera CVW material.