Liqid’s had a great run of late with some eye-popping deals to provide composable supercomputing resources for artificial intelligence operations by the Department of Defense. In the past week, we were lucky to be able to announce an agreement to provide a 17-petaflop composable disaggregated system for the US Army Engineering and Support Center (ERDC), Huntsville, Alabama, and two systems totalling 15 petaflops of compute power for the U.S. Army Combat Capabilities Development Command’s Army Research Laboratory’s DoD Supercomputing Resource Center at Aberdeen Proving Ground in Maryland. At GTC Fall 2020 this year, we made an announcement highlighting the central role of the powerful NVIDIA A100 GPUs and NVIDIA Mellanox HDR 200 gigabit per second InfiniBand smart networking to all three systems.
Once completed, the systems will be the largest composable supercomputers in deployment. At the time of this publication, in fact, the Huntsville deployment would register at No. 15 on the TOP500 ranking of the world’s most powerful high-performance computing (HPC) platforms.
In a recent HPCwire article, Tiffany Trader spoke with George Moncrief, chief technologist at ERDC’s DoD Supercomputing Resource Center (DSRC), who expanded on the decision-making process behind going with Liqid’s solutions.
“We try to have a balanced approach to the evaluation; we’d like a good price-performance number as well as a good fit for our requirements,” said Moncrief. “And we are also looking at maximizing competition. So we try and write these requirements in such a way and evaluate all of those factors, so that as many vendors as possible have an opportunity to win.”
For each of these important wins Liqid beat out some of the industry’s most well-established data center players by delivering composable supercomputing solutions capable of right-sizing resources to meet shifting data demand. With Liqid’s Composable Fabric, users can dynamically orchestrate any CPU-to-GPU ratio, and incorporate other accelerators across PCI-Express Gen 4.0, Infiniband, and Ethernet fabrics for a perfectly balanced system.
Liqid’s orchestration software dynamically composes GPUs, NVMe SSDs, networking, and storage-class memory to create software-defined bare-metal servers on demand. This enables unparalleled resource utilization to deliver previously impossible performance for AI-driven data analytics operations.
Registered GTC Fall attendees can view Liqid’s presentation here. To learn more about Liqid, go to www.liqid.com. To learn more about Liqid’s composable disaggregated infrastructure technology, contact us to discuss your specific data center needs with a representative. Follow Liqid on Twitter and LinkedIn to stay up to date with the latest Liqid news and industry insights.
