TACC’s Lonestar6 supercomputer gets GPU and Server Boost

DALLAS, November 14, 2022 – The Texas Advanced Computing Center (TACC) today announced at SC22 in Dallas that the Lonestar6 supercomputer, which entered full production in January 2022, has just received a boost of new servers and GPUs from Dell Technologies and NVIDIA. This system for Texas researchers allows them to compute and compete at the cutting edge of science and engineering.

TACC Executive Director Dan Stanzione stands next to Lonestar6 in the data center. The system was updated with three new Dell R750 server racks with a total of 240 NVIDIA A100 40 GB Tensor Core GPUs interconnected by the NVIDIA Quantum 200 Gb/s InfiniBand network. Credit: TACC.

“Lonestar6 replaces Longhorn as the largest GPU resource in TACC’s ecosystem and the primary platform for the growing application of AI among our users,” said Dan Stanzione, TACC Executive Director and Associate Vice President. for research at the University of Texas at Austin. “With the Lonestar6 upgrade, we continue to meet the growing demand for AI and other GPU-accelerated problems, and leverage the power efficiency of heterogeneous computing.”

The system has been updated with three new Dell R750 server racks with a total of 240 NVIDIA A100 40 GB Tensor Core GPUs interconnected by NVIDIA Quantum 200 Gb/s InfiniBand networking, the only fully networked computing platform dischargeable in the world.

“Lonestar6 is an example of the innovation and opportunity supercomputing offers to solve society’s most complex challenges while pushing the boundaries of efficiency for accelerated computing,” said Ian Buck, vice -President of Hyperscale and High Performance Computing at NVIDIA. “NVIDIA Accelerated Computing equips next-generation systems like the Lonestar6 with the extreme performance and efficiency to enable AI innovation for the scientific community.”

Upgrades to Lonestar6 add around five quadrillion math operations per second – or, in high-performance computing (HPC) terminology, five petaflops of computing power. This is in addition to the existing three petaflops of performance from the system’s AMD processors.

A person would have to make a calculation every second for 150 million years to match what Lonestar6 will calculate in just one second. It will allow doctors to design patient-specific cancer treatments, astronomers to peer deeper into the cosmos than ever before, and help meteorologists predict how our climate is changing.

More recently, researchers relied on Lonestar6 to simulate Hurricane Ian for 10 days around the clock to analyze storm surge, flooding, wind and wave information from ocean models. coastal and inland flooding.

More than 90,000 unique users have viewed data on the Coastal Emergency Risk Assessment website, including emergency managers, first responders, and experts from FEMA, the U.S. Department of Transportation, and many other agencies.

Lonestar6 is funded by the University of Texas Research Cyberinfrastructure, a collaboration between TACC and the University of Texas System, and is supported by several academic institutions in Texas.

Texas’ support of advanced computing through the initiative has made the state a national leader in computer science and engineering over the past decade, with research benefits helping to impact on citizens, industry and society.

Stampede2 extended through June 2023

TACC also announced earlier this year that the National Science Foundation (NSF) has extended operational funding for the Stampede2 supercomputer through June 2023.

As part of the expansion, 448 of the oldest Intel Knights Landing Xeon Phi nodes were upgraded with 224 of the latest generation of two-socket Dell PowerEdge R650 servers using Intel Xeon Platinum 8380 “Ice Lake” processors. These latest servers will provide more than double the performance and nearly triple the memory of previous nodes to provide additional capabilities to the Stampede2 system.

“Stampede2 has now run over 10 million jobs, and despite its advanced age, it still has the most users of any system we manage, supporting capacity computing and a broad scientific mission,” said said Stanzione. “We felt it was important not only to extend the lifespan, but also to update the hardware to continue to support cutting-edge science.”

Stampede2 has over 20 petaflops of overall performance, still making it the largest system by delivery time in the NSF-funded ACCESS program, despite being rolled out in 2017.

“We imagine the new IceLake nodes will continue to work in some systems long after Stampede2 retires next year,” Stanzione said.

Frontera and the Leadership Class Computing Facility (LCCF)

TACC’s Frontera supercomputer will operate until 2025 when the systems of the new Leader Class Computing Facility (LCCF), led by a new machine named “Horizon”, are expected to come online.

TACC will work with new partners to build a bespoke colocation facility to house the core systems of LCCF, the center’s first off-campus enterprise. The LCCF will be hosted in a commercial Switch data center on the Round Rock, Texas campus. Credit: TACC

TACC announced this year that the center will work with new partners to build a bespoke colocation facility to house the core systems of LCCF, the center’s first off-campus enterprise. The system will be housed in a commercial Switch data center on the Round Rock, Texas campus. This new space will add over 15 MW of capacity, bringing TACC’s total data center footprint to 25 megawatts of available power.

The LCCF will launch a new era of computational discovery, including a system that can produce 10 times more science than Frontera, and partnerships with leading HPC organizations across the country.

Part of the 10x improvement will come from code performance optimizations. TACC announced earlier this year the set of 21 “Grand Challenge” class science codes and problems that will receive funding under the “Characteristic Science Applications” program.

Applications include software for major international experiments such as the IceCube neutrino observatory; codes widely used by seismic and astrophysical communities; and custom codes that explore new approaches to machine learning and black hole modeling.


Source: Faith Singer, TACC

About Mariel Baker

Check Also

The host city’s internet-based economy is growing rapidly

An aerial view of Wuzhen, the permanent site of the World Internet Conference. [Photo/Xinhua] As …