During SC16, AMD announced that the Radeon GPU technology will be available to Google Cloud Platform users in 2017. As a result, Google will deploy AMD’s fastest single-precision dual GPU compute accelerators and Radeon-based AMD FirePro S9300 x2 Server GPUs to accelerate both Google’s Compute Engine and Cloud Machine Learning services. Rather impressively, AMD’s FirePro S9300 x2 can cope with highly parallel calculations including medical and financial simulations, seismic and subsurface exploration, machine learning, video rendering, video encoding and scientific analysis. Raja Koduri, senior vice president and chief architect, Radeon Technologies Group, AMD said about this new move:
“Graphics processors represent the best combination of performance and programmability for existing and emerging big data applications,”
“The adoption of AMD GPU technology in Google Cloud Platform is a validation of the progress AMD has made in GPU hardware and our Radeon Open Compute Platform, which is the only fully open source hyperscale GPU compute platform in the world today. We expect that our momentum in GPU computing will continue to accelerate with future hardware and software releases and advances in the ecosystem of middleware and libraries.”
On another note, AMD recently revealed the Radeon Open Compute Platform (ROCm) to speed up development of high-performance, energy-efficient heterogeneous computing systems. The company is currently running demonstrations of a wide range of GPU advances including:
- ROCm Technology Cluster running Machine Learning Code on Supermicro servers
- Porting the CUDA application Caffe via HIP Porting Tool
- Ray-tracing and VR visualisation for HPC with AMD FirePro S9300 X2 & Radeon R9 Nano GPUs
- OpenMP 4.5 Interoperability targeting multiple GPUs & platforms
- IBM Power8 server with AMD FirePro S9170 Server GPU running ROCm
- Penguin Computing Tundra Extreme ARMv8 ThunderX based server with Radeon RX 460 running ROCm
- In-situ rendering with Headless OpenGL/EGL Interop OpenCL on ROCm
