Amd announces support for tensorflow in rocm and improves epyc availability

AMD has been steadily increasing the production and availability of its EPYC processors, its new 32-core, 64-thread monsters based on the Zen architecture that stand out for offering a better feature set in single-socket configurations than even some two-socket configurations. sockets from Intel. In addition, support for TensorFlow in ROCm has been added.

AMD bets on TensorFlow

Now AMD announces a greater availability of solutions powered by its EPYC processors and among which we can find OEMs, distributors and system integrators such as ASUS, BOXX, GIGABYTE, HPE (Hewlett Packard Enterprise), Penguin Computing, Supermicro and Tyan. ASUS bets heavily on HPC and virtualization through its RS720A-E9 and RS700A-E9. BOXX chose to combine EPYC CPUs with Radeon Instinct accelerators for multi-GPU computing and deep learning solutions. GIGABYTE is moving forward with rackmount servers, and Supermicro is moving from its well-known tower form factors to 1U, 2U, and 4U for HPC and storage.

AMD Ryzen Threadripper 1950X & AMD Ryzen Threadripper 1920X Review in Spanish (Analysis)

On the other hand, we cannot forget the importance of software to be able to get all the benefits of hardware, which is why AMD has doubled its efforts in the platform for high-performance computing and open standards ROCm (Radeon Open Compute platforM) that reaches the version 1.7. This new version of ROCm has added support for the TensorFlow and Caffe machine learning frameworks in MIOpen libraries.

TensorFlow is gaining vital importance in the artificial intelligence sector so all companies are strongly betting on it, Nvidia with its Volta architecture and Google is not far behind either.

Techpowerup font