NVIDIA and Azure users can now access 35 GPU-accelerated containers for HPC applications
NVIDIA GPU Cloud (NGP), often used by AI researchers and data scientists, is now supported in the cloud by Microsoft Azure.
NGP is made up of pre-integrated, ready to run, GPU-accelerated containers that let developers and researchers control how they scale their high computational workloads.
With high performance computing (HPC) deploying and maintaining the latest software in clusters of systems is an ongoing concern. As applications and frameworks are updated constantly, building a project with one setting in mind is a difficult task.
Commenting in a blog Chris Kawalek product marketing manager at NVIDIA GPU Cloud stated that: “Building and testing reliable software stacks to run popular deep learning software such as TensorFlow, Microsoft Cognitive Toolkit, PyTorch and NVIDIA TensorRT is challenging and time consuming.”
“There are dependencies at the operating system level and with drivers, libraries and runtimes. And many packages recommend differing versions of the supporting components.”
“We test, tune and optimize the complete software stack in the deep learning containers with monthly updates to ensure the best possible performance.”
NVIDIA GPU Cloud on Azure
With the cross support between NVIDIA and Azure, users can now access 35 GPU-accelerated containers for HPC applications, deep learning software, and HPC visualisations tools. All of which will run on certain Azure instance types with NVIDIA GPU’s. Containers from the NGC container registry are now supported on VVIDIA Volta and Pasca-Powered Azure ND, NCv2 and NCv3.
Microsoft Azure meanwhile has made a general release of the Azure CycleCloud which is a tool designed to operate and optimise HPC clusters. Clusters often refer to a group of computers closely working together. The aim of CycleCloud is to make HPC burst, cloud and hybrid clusters deployable in a controllable and scalable manner.
HPC is been used by an array of industries to undertake heavy computational tasks and research. In the pharmaceutical sector it is been utilised to assess how proteins fold and what structures they make, which helps to enable better drug designs.
Johnson & Johnson use Azure CycleCloud and HPC Clusters to simplify management of data workloads and control access and costs.
Brett Tanzer Partner PM Manager at Azure Specialized Computer commented in a blog that: “HPC IT administrators can deploy high-performance clusters of compute, storage, filesystem, and application capability in Azure.”
“Azure CycleCloud’s role-based policies and governance features make it easy for their organizations to deliver the hybrid compute power where needed while avoiding runaway costs. Users can rely on Azure CycleCloud to orchestrate their job and data workflows across these clusters.”