STAC Report: STAC-A2 (derivatives risk) on 8 x NVIDIA A100 80GB GPUs and Red Hat OpenShift

Containerized NVIDIA A100 GPUs set 8 records.

9 November 2021

STAC recently performed STAC-A2 Benchmark tests on a solution using eight of the latest NVIDIA A100 GPUs and a major update to the STAC-A2 Pack for CUDA. These are also the first published STAC-A2 results for a solution using containers.

The STAC Report is now available here.

STAC-A2 is the technology benchmark standard based on financial market risk analysis. Designed by quants and technologists from some of the world's largest banks, STAC-A2 reports the performance, scaling, quality, and resource efficiency of any technology stack that is able to handle the workload (Monte Carlo estimation of Heston-based Greeks for a path-dependent, multi-asset option with early exercise).

The stack consisted of the STAC-A2 Pack for CUDA (Rev G) using CUDA 11.2 on 8 x NVIDIA A100 (Ampere) SXM4 80GB GPUs in a DGX A100 server running Red Hat^® OpenShift^® 4.8.3.

NVIDIA and Red Hat wished to highlight several results from this report:

• Compared to all publicly reported solutions to date, this containerized GPU-based solution set 8 records:
• the highest energy efficiency (STAC-A2.β2.HPORTFOLIO.ENERG_EFF)
• the highest space efficiency (STAC-A2.β2.HPORTFOLIO.SPACE_EFF)
• the highest throughput (STAC-A2.β2.HPORTFOLIO.SPEED)
• the fastest warm and cold times in the large Greeks benchmark (STAC-A2.β2.GREEKS.TIME.{WARM,COLD})
• the fastest warm time in the baseline Greeks benchmark (STAC-A2.β2.GREEKS.TIME.WARM)
• the highest maximum paths (STAC-A2.β2.GREEKS.MAX_PATHS)
• the highest maximum assets (STAC-A2.β2.GREEKS.MAX_ASSETS)
• Compared to the previous best results for a coprocessor-based solution (SUT ID NVDA200909), this solution had:
• 3.0x the throughput (STAC-A2.β2.HPORTFOLIO.SPEED vs. SUT ID INTC210331)
• 2.6x the energy efficiency (STAC-A2.β2.HPORTFOLIO.ENERGY_EFF vs. SUT ID INTC210315)
• 2.6x the speed in the warm baseline Greeks benchmark (STAC-A2.β2.GREEKS.WARM vs. SUT ID NEC210422)
• 2.3x the speed in the warm large Greeks benchmark
  (STAC-A2.β2.GREEKS.10-100k-1260.TIME.WARM vs. SUT ID INTC181012)
• 2.1x the maximum assets (STAC-A2.β2.GREEKS.MAX_ASSETS vs. SUT ID INTC181012)

For details, please see the report at the link above. Premium subscribers have access to the code used in this project as well as the micro-detailed configuration information for the solution. To learn about subscription options, please contact us.