Overview of the Intel® Optimized HPL-AI* Benchmark
The Intel® Optimized HPL-AI Benchmark is heavily modified based on the High-Performance LINPACK (HPL) Benchmark (http://www.netlib.org/benchmark/hpl/) from Innovative Computing Laboratories (ICL) at the University of Tennessee, Knoxville to meet the requirements of the HPL-AI (also known as HPL-MxP) rules (see https://hpl-mxp.org/rules.md). The usage of the Intel® Optimized HPL-AI Benchmark is very similar to that of the Intel® Distribution for LINPACK Benchmark. Although the Intel® Optimized HPL-AI Benchmark measures the same amount of time as the Intel® Distribution for LINPACK Benchmark, lower precision is used to obtain an approximate solution and the final solution is calculated by a double-precision–based iterative method.
For more details about the HPL-AI (HPL-MxP) benchmark, see the HPL-MxP home page.
Product and Performance Information
Performance varies by use, configuration and other factors. Learn more at www.Intel.com/PerformanceIndex.
Notice revision #20201201