In today’s world of big computing and even bigger data, many HPC users are taking advantage of GPU-based clusters to solve scientific and engineering challenges. In a heterogeneous computing model with CPUs and GPUs working together, the sequential part of the application runs on the CPU and the computationally intensive part runs on the GPU. By exploiting the massive parallelism in GPUs, users can run applications substantially faster than they could on a CPU-only system — especially data-intensive science and analytics applications.
The GPU is also a keystone for today’s artificial intelligence (AI). Because of its large memory capacity and superior computational ability, GPU-accelerated systems beat CPU-only systems on data-intensive workloads like machine learning and autonomous vehicle design.
Altair PBS Works™ supports GPU scheduling so users can take advantage of these powerful architectures to accelerate processing.
“At JAIST, it is imperative to provide our world-class scientists with the best available technology resources... The pioneering Cray XC supercomputer with Altair PBS Professional [running NVIDIA® Tesla® GPU accelerators] will allow our users to expand the scope of their research efforts with a proven, well-integrated solution they can rely on.”
—Director of Research Center for Advanced Computing Infrastructure at the Japan Advanced Institute of Science and Technology (JAIST)
Altair PBS Professional, part of the Altair PBS Works suite, supports advanced GPU scheduling, the ability for a job to separately allocate each individual GPU on a node. That means you can share a single node among multiple jobs where each job requires its own GPUs.