NVIDIA at Hot Chips 2026
NVIDIA joined Hot Chips 2026 at Stanford University to showcase how the NVIDIA Accelerated Computing platform is reimagining the data center for the age of AI. Featuring next-generation AI infrastructure—the NVIDIA Vera CPU, NVIDIA Rubin GPU, NVLink™ Switch, and LPU architecture, paired with Spectrum-X™ Ethernet networking and BlueField™-4 DPUs—and a rich software stack, it's built to take on the modern AI factories' toughest challenges.
For more information on the 2026 sessions, visit the Hot Chips webpage.
News at Hot Chips 2026
Keynote Address by Gilad Shainer at Hot Interconnects 2026
Don’t miss NVIDIA SVP of Networking Gilad Shainer on the keynote stage at Hot Interconnects 2026. In “Networking Innovations for Gigascale AI Systems,” we’ll explore how the network has become the defining architecture of gigascale AI as infrastructure scales to hundreds of thousands, and soon millions, of interconnected GPUs. Expect a deep dive into the innovations enabling gigascale infrastructure to train massive foundation models, power real-time inference pipelines, and interconnect federated AI factories globally.
2026 Schedule at a Glance
Sunday, August 23 | 3:45 p.m.
RISC-V Profile and Platform for Interoperability With NVIDIA GPUs
Tutorial | Frans Sijstermans, NVIDIA
Monday, August 24 | 10:30–11 a.m.
NVIDIA's Vera CPU
Talk | Jonathon Evans and Polychronis Xekalakis, NVIDIA
Monday, August 24 | 4:45–5:15 p.m.
NVIDIA Rubin GPU: Driving the Era of Agentic AI
Talk | Manas Mandal, Rajballav Dash, and Ravi Manyam, NVIDIA
Tuesday, August 25 | 11:30 a.m.–noon
NVIDIA BlueField-4 Processor Powers the AI Factory Operating System
Talk | Idan Burstein, NVIDIA
Tuesday, August 25 | noon–12:30 p.m.
NVIDIA Spectrum-X Multiplane Network Architecture
Talk | Gilad Shainer, NVIDIA
Tuesday, August 25 | 2:45–3:15 p.m.
Think Fast: LPU Accelerator for Heterogeneous Compute
Talk | Igor Arsovski and Santosh Raghavan, NVIDIA