Speaker
Description
Memory and storage read and write costs can lead to a significant loss of time and energy in current HPC systems. Byte-addressable non-volatile memory (NVM) could provide considerable improvements in both time and energy requirements over conventional DRAM memory. Using Optane DCPMM, Intel's new byte-addressable and persistent memory, the NEXTGenIO project investigated the performance of NVRAM by designing, building and testing a bespoke prototype NVM system. The main goal of the project was to explore the potential of NVRAM in overcoming performance bottlenecks in I/O and main memory, which are considered significant barriers to Exascale computing.
In this talk we will give a brief overview of the NEXTGenIO system (192GB DRAM and 3TB of NVM per dual socket node), and the various NVRAM usage modes. The results from a number of investigative test cases run on the NEXTGenIO prototype system will be presented. In particular we will discuss I/O performance, run-time, and energy consumption for applications with large I/O demands, such as OpenFOAM and CASTEP. Comparison of the results from NVRAM and DRAM shows that NVRAM can indeed provide significant improvement in both performance and energy consumption.
