The Path to Embedded Exascale
Session ChairWu Feng
Event Type
Emerging Technologies
Accelerators
Energy
Exascale
TimeWednesday, November 16th10am - 6pm
Location155-B
DescriptionIn June 2016, NVidia released JetPack 2.2 for the ARMv8 powered Jetson TX1. The release of unified 64 bit kernel, userspace, and CUDA 7.5 libraries significantly increased performance per watt over the previous JetPack, which was limited to 32-bits. Demonstrations on production codes and traditional benchmarks have shown the JTX1 to be on an aggressive path that will put us back on a Moore’s Law trajectory as we approach the exascale era. The interplay of ARM commands, NEON, and UMA-enabled CUDA code has drastically increased embedded performance relative to their discrete analog. The authors have experimented with other TX1-based products which have had an unsupported 64-bit userspace since shortly after they shipped. Demonstrations of the latest prototypes will accompany the presentation.
