The High Performance LINPACK (HPL) benchmark is a software package that solves a (randomly generated) dense linear system in double precision (64-bit) arithmetic on distributed-memory computers. Written in a portable ANSI C and requiring an MPI implementation as well as either the BLAS or VSIPL library, HPL is often one of the first programs to run on large computer installations, producing a result that can be submitted to the biannual TOP500 list of the world’s fastest supercomputers.
HPL 2.2, released in 2016, includes several bug fixes and accuracy enhancements based on user feedback. The major focus of HPL 2.2 is to improve the accuracy of reported benchmark results, and ensure scalability of the code on large supercomputer installations with hundreds of thousands of computational cores. The last version also featured a detailed time-of-run accounting to help with assessing power requirements at the time of execution, a metric which has been reported with TOP500 results since 2007 and is also highlighted on the Green500 list. In 2011, the LINPACK benchmark app for iOS achieved performance of over 1 Gflop/s on an Apple iPad 2, with per-Watt performance easily beating supercomputing solutions, including the most power-efficient systems based on hardware accelerators. The App now achieves over 4 Gflop/s on the iPad Air and Pro.
Find out more at http://icl.utk.edu/hpl/
In Collaboration With
- Oak Ridge National Laboratory