Read more at the NVIDIA Blog:
“Scaling” of software applications is, in many cases, the only way to model complex physical phenomena on high-performance computing (HPC) systems. But, today’s CPU-based clusters aren’t getting the job done. The HPC industry hasn’t made much progress in this area, even though some of the smartest developers are working on HPC applications. This is because programming to scale out applications is really difficult.
However, HPC customers are successfully scaling their scientific applications to thousands of GPUs (graphics processing units). For example, researchers at the Chinese Academy of Sciences Institute of Process Engineering (CAS-IPE) scaled their research to develop more efficient solar panel technologies to thousands of GPUs. Likewise, researchers in France and Japan aim to provide a better understanding of earthquakes are accelerating their science on a grand scale with GPUs. This video provides additional examples of application scaling on GPU-based HPC systems.
