Nvidia Announces Support For Arm CPUs: Why It Matters

NVIDIA Corporation NVDA announced Monday at the International Supercomputing Conference that it will now support Arm CPUs. 

What Happened

Nvidia's full stack of AI and HPC software will be made available to the Arm ecosystem by the end of 2019, the company said.

The stack will include all CUDA-X AI and HPC libraries, GPU-accelerated AI frameworks and software development tools like PGI compilers with OpenACC support and profilers.

The announcement is the result of 10 years of collaboration between Nvidia and Arm-based HPC systems. Nvidia uses Arm for some of its own systems in its portable gaming, self-driving technologies, robots and embedded AI computing, the company said.

Why It's Important

Combining Nvidia's CUDA-accelerated computing with Arm's CPU architecture will "give the HPC community a boost to exascale," Nvidia CEO and founder Jensen Huang said in a statement.

The technology to power supercomputers represents the potential to expand "the frontier of human knowledge," he said. 

Arm CEO Simon Segars said the company "is working with our ecosystem to deliver unprecedented compute performance gains and exascale-class capabilities to Arm-based SoCs." 

What's Next

Several global experts and leaders in the HPC industry and Arm ecosystem commented on the latest development in Nvidia's press release with a great degree of enthusiasm.

Ampere Computing chairman and CEO Renee James said Nvidia's announcement will "accelerate our work in building out the software ecosystem for Arm-based servers and enable breakthrough Ampere platforms with NVIDIA GPUs for efficiency and performance."

Thomas Lippert, the director of the Jülich Supercomputing Centre, said the Nvidia news is an exciting development "which is essential for the establishment of true modularity for supercomputers and composable data centers of the future." 

Nvidia shares were up 0.9% at $145.90 at the time of publication Monday. 

Related Links:

Analyst: AMD's New Graphics Cards Make 'Meaningful Improvement' But Cost More Than Expected

Oppenheimer Bullish On Nvidia, Cautious On AMD, Intel