Texas Advanced Computing Center (TACC) on Monday announced Stampede3 is a new supercomputer used for high-end simulation and artificial intelligence/machine learning applications. The new system is based on Intel’s Xeon CPU Max processors with on-package his HBM2E memory and Data Center Max computing GPUs.
“We will continue our long-standing partnership with Dell and Intel to leverage NSF’s investment in Stampede2 for this new scientific resource that uses state-of-the-art processors with high-bandwidth memory and makes Intel’s graphics processing units widely available to the NSF open science community,” said Dan Stanzione, Executive Director of TACC.
TACC’s Stampede3 is built by Dell using the latest hardware from Intel. The new machine consists of 560 nodes based on 56-core Intel Xeon CPU Max processors with 64 GB of on-package HBM2E memory, delivering about 63,000 general purpose cores and peak performance of about 4 FP64 petaFLOPS. Additionally, Stampede 3 includes his 10 Dell PowerEdge XE9640 servers with 40 new Intel Data Center GPU Max computing GPUs, codenamed Ponte Vecchio, for AI/ML capabilities.
“We believe the high-bandwidth memory of the Xeon Max CPU nodes will give users better performance than any CPU they have ever seen,” said Stanzione. “Offers more than double the memory bandwidth performance per core compared to current 2nd and 3rd Gen Intel Xeon Scalable processor nodes on Stampede2.”
The new supercomputer uses the latest Omni-Path Fabric 400 Gb/s technology and features a backplane bandwidth of 24 TB/s, enabling great scalability and low latency for various applications requiring high-precision simulation.
In addition to Intel’s latest hardware, Stampede3 plans to reintegrate Stampede2 nodes based on previous generation Xeon Scalable CPUs for high memory applications, high throughput computing, interactive workloads and other smaller workloads.
According to TACC, the new system will use 1,858 compute nodes with over 140,000 cores, over 330TB of RAM, 13PB of new storage, and peak performance of nearly 10 petaflops.
Like its predecessors Stampede in 2012 and Stampede2 in 2017, Stampede 3 will be an integral part of the National Science Foundation (NSF). access A scientific supercomputing ecosystem, used for open science research projects.
TACC expects Stampede3 to be available in the fall of 2023, followed by full-scale operation in early 2024. From 2024 he will serve the open science community until 2029.
