End-to-End Support Toward Accelerating Discovery and Innovation with a Converged HPC and AI Supercomputer
A workload-driven system capable of running HPC/AI workloads is more important than ever. Organizations face many challenges when building a system capable of running HPC and AI workloads. There are also many complexities in system design and integration. Building a workload driven solution requires expertise and domain knowledge that organizational staff may not possess.
This paper describes how Quanta Cloud Technology (QCT), a long-time Intel® partner, developed the Taiwania 2 and Taiwania 3 supercomputers to meet the research needs of the Taiwan’s academic, industrial, and enterprise users. The Taiwan National Center for High-Performance Computing (NCHC) selected QCT for their expertise in building HPC/AI supercomputers and providing worldwide end-to-end support for solutions from system design, through integration, benchmarking and installation for end users and system integrators to ensure customer success.
QCT meets customer needs for creating HPC/AI systems with demand analysis evaluation, architecture design, system deployment, system tuning/ benchmarking, and pilot run and implementation:
• QCT provides rapid deployment kits to quickly set up its operating system and HPC/AI environments
• Web-based UI allows administrators to monitor and manage the cluster easily
• QCT Orqestra software allows out-of-band system monitoring for up to 5000 nodes with control by BMC
• QCT AI Labs can do proof-of-concept, testing, tuning, and benchmarking for end users or ISV partners
• QCT has HPC/AI expertise for different scales, different architectures, and domain expertise on benchmarking and tuning
• QCT provides end-to-end support for end users and system integrators to ensure customer success
