In this 30-minute RC Quick Byte*, we will provide an overview of the cutting-edge capabilities of NVIDIA's Grace Hopper (GH200) Superchip, including how to request access and deploy software on the GH200s hosted by CU Research Computing. The GH200 is designed to revolutionize high-performance computing by providing shared memory between the CPU and GPU. This design can be useful for a large range of data-intensive applications and has proven fruitful for Machine Learning and Artificial Intelligence workloads. This introductory talk is useful for anyone curious about recent advancements in computing or individuals interested in using Research Computing’s GH200 nodes.
- _What’s an RC Quick Byte?: An RC Quick Byte is a 20-30 minute sessions intended to provide essential information about a specific topic. You can expect to leave a 'Quick Byte' with the information you need to apply a tool, service, or concept.