Liqid, provider of the world's most comprehensive composable disaggregated infrastructure (CDI) platform, will return this year to Supercomputing 2021, where the company will demonstrate how its Liqid Matrix CDI software is helping to redefine high-performance computing (HPC) for the age of artificial intelligence and machine learning (AI+ML). In booth #3001, Liqid will demonstrate its capability to compose for NVIDIA A100 Tensor Core GPUs in AI-driven HPC environments, as well as highlight new customers and research, and offer an early look at new deployment models for CDI. Liqid will also showcase its award-winning CDI software platform for AI and HPC with channel partners at Carahsoft (#1850) and One Stop Systems (#3019).

"We're excited to be back at SC21 this year to bring Liqid Matrix CDI software to organizations that are eager to move past the static limitations of legacy HPC configurations and deliver a dynamic, software-defined platform for material breakthroughs in academic and scientific research," said Sumit Puri, CEO & co-founder, Liqid. "Taking full advantage of high-performing NVIDIA A100 GPUs for AI workloads, we plan to enable more rapid deployment of GPU resources with Liqid Matrix CDI software that can dynamically adapt for modern data center-scale workloads as required, and scale beyond static HPC architectures to meet tomorrow's challenges."

Liqid Software with NVIDIA A100 Tensor Core GPUs: CDI for the World's Toughest Problems

Liqid Matrix software delivers disaggregated composability for NVIDIA A100 Tensor Core GPUs with increased utilization, flexibility, and time to value for AI+ML, HPC, and other high-value applications driving global economic development.

NVIDIA A100 GPUs deliver unprecedented acceleration at every scale to power the world's highest-performing elastic data centers for AI, data analytics, and HPC. Powered by the NVIDIA Ampere architecture, A100 GPUs are the engine of the NVIDIA data center platform. Providing up to 20x higher performance over the prior generation, they can be partitioned into seven GPU instances to dynamically adjust to shifting demands.

Liqid, an Elite member of the NVIDIA Partner Network (NPN) program, delivers improved resource management and utilization with disaggregated infrastructure, featuring software-defined infrastructure that offers full, granular composability for NVIDIA A100 GPU-based solutions. Composable, high-performance NVIDIA A100 GPU resources are allocated through software at bare metal across Liqid's ultra-low-latency platform, simplifying scalability and resource orchestration.

Unlocking the performance and efficiency necessary for a wide variety of use cases in AI+ML and HPC, Liqid enables the clustering of dozens of NVIDIA A100 GPUs for the most demanding workloads, real-time GPU-to-CPU allocation, and even CPU bypass for "peer-to-peer" data transactions. In addition, Slurm and VMware vCenter integration provides transparent project management for AI in HPC and enterprise environments, delivering the most comprehensive composable disaggregated software platform for AI and HPC.

"The NVIDIA-accelerated computing platform features A100 GPUs, which provide the highest performance for HPC and AI applications to address the most pressing computational challenges in industry and science," said Paresh Kharya, senior director of product management and marketing for accelerated computing, NVIDIA.