NVIDIA Developments for CFD Applications on GPUs and Arm CPUs

Speaker: Stan Posey & John Linford, NVIDIA
September 7, 2023

Presentation

Abstract

Recent NVIDIA developments in GPU-based HPC and AI have substantially improved computational speeds for a broad range of CFD software relevant to engineering and scientific practice. The first topic of this talk will examine some of the latest CFD implementations for GPUs with examples from commercial and research application software. GPU programming strategies of choice can depend on several factors such as availability of suitable GPU-based libraries including linear solver level, distributed parallel method, performance portability requirements for cross-platform deployment, and execution profile characteristics among others. The second topic will introduce a new data center system design for power-efficient HPC based on the NVIDIA Grace Arm CPU and latest H100 GPU. The NVIDIA Grace CPU offers 1 TB/s of memory bandwidth, and communicates at data transfers speeds of up to 900 GB/s bidirectional bandwidth with another Grace CPU or the NVIDIA Hopper H100 GPU that offers 3 TB/s of memory bandwidth. The NVIDIA Grace Hopper software environment will also be described and include requirements for build and execution procedures of CFD software, and examples of performance comparisons with the latest x86 platforms.

Biography

Stan Posey joined NVIDIA in 2009 with more than 20 years of experience in the HPC industry. Posey began his career as an engineering analyst in applied CFD at the U.S. DOE Oak Ridge National Laboratory, and mechanical CAE consultant CD-adapco, and later in technical roles at system vendors Silicon Graphics and Panasas. At NVIDIA, Posey is a program manager who leads the company’s strategy on HPC solutions for the domains of CFD and Earth system science. Posey graduated with M.Sc. and B.Sc. degrees in Mechanical Engineering from the University of Tennessee Knoxville, and is a member of the AIAA who serves on the AIAA Computer Systems Technical Committee.

John Linford is focused on developing the NVIDIA CPU software ecosystem for supercomputing and the datacenter. He previously worked at Arm as a director of HPC engineering. John has almost two decades of front-line experience with HPC applications, systems, and optimization experience and particularly enjoys working with emerging technologies and extreme-scale systems. He received his Ph.D. in computer science from Virginia Tech in 2010 and lives in Austin, Texas.

Seminar Participation

Attending In Person

For select seminars, you may wish to attend in-person if you are in the vicinity of NASA Ames Research Center. These AMS seminars are held in Building N258 in Room 127, just off the front lobby. Sessions available to attend in-person will be indicated in the announcement information. In-person participants without NASA badges will need a valid visitor badge (US citizenship or permanent residency is required) to enter the facility.

Join Us Online

To attend an AMS seminar, click on the Microsoft Teams link in the seminar description. You can connect through the desktop Teams App if you have it, or you can select "Watch on the Web Instead" to enter the meeting as a guest. Please ensure that you've muted your microphone in Teams before you select "Join Now" to enter the meeting lobby.