Course Overview
The course Run Intel Tools in the Cloud: Intel AMX & Intel AVX-512 Demonstration is designed to help users understand how to take advantage of hardware optimizations to get optimal AI model performance. The course focuses on the difference between performance with and without Intel AVX-512 and evaluates the performance difference between Intel AVX-512 and VNNI, as well as the difference between Intel AVX-512 and AMX. Through a demonstration, users will learn how to benchmark AI models using different instructions and models, including FP32, int8, and AVX-512. The course covers the use of Intel Model Zoo, a GitHub repository that provides optimized deep learning models, and shows how to use NUMA control to bind workloads to physical cores. By the end of the course, users will be able to set a baseline on a system, enable higher instruction sets, and compare performance to show the additional performance gained with those instruction sets.
Share With
Chapter Overview
The course Run Intel Tools in the Cloud: Intel AMX & Intel AVX-512 Demonstration is designed to help users understand how to take advantage of hardware optimizations to get optimal AI model performance. The course focuses on the difference between performance with and without Intel AVX-512 and evaluates the performance difference between Intel AVX-512 and VNNI, as well as the difference between Intel AVX-512 and AMX. Through a demonstration, users will learn how to benchmark AI models using different instructions and models, including FP32, int8, and AVX-512. The course covers the use of Intel Model Zoo, a GitHub repository that provides optimized deep learning models, and shows how to use NUMA control to bind workloads to physical cores. By the end of the course, users will be able to set a baseline on a system, enable higher instruction sets, and compare performance to show the additional performance gained with those instruction sets.