Skip to main content
We are proud to announce the general availability of NVIDIA HGX B200-based instances on CoreWeave. These instances deliver exceptional performance for AI training and real-time inference in enterprise AI workloads, offering upwards of 2x faster performance per GPU compared to previous-generation NVIDIA Hopper GPUs.

​
Overview

Powered by eight NVIDIA B200 Blackwell GPUs, each with 180 GB of HBM3e memory, these instances are designed to meet the demands of modern AI workloads. Networking is supported by an NVIDIA BlueField-3 DPU and eight NVIDIA ConnectX-7 InfiniBand host channel adapters (HCAs), delivering high throughput and lossless scaling. A 400G NDR non-blocking NVIDIA Quantum-2 InfiniBand fabric ensures seamless, high-speed connectivity across nodes. Contact our Sales team to learn more about NVIDIA B200 instances and their capabilities.

​
Additional resources

Last modified on April 2, 2026