Powering the Future of Innovation: How CoreWeave Is Advancing AI Cloud Performance With BlueField-4

Powering AI Innovation: CoreWeave + NVIDIA BlueField-4 | CoreWeave Blog

Announcement

Announcement

Webinar

Announcement

Podcast

Announcement

GTC 2026

Announcement

CoreWeave brings up the industry’s first NVIDIA Vera Rubin NVL72 deployment.

Read more

Products

Data and storage

Infrastructure control

Runtime acceleration

Model and agent development

Mission control

Solutions

Pricing

Resources

About us

Contact us Login

Contact us Login

Clear

The generative AI revolution isn’t coming — it’s already here and  actively transforming every industry around the world. As models grow larger and workloads more complex, traditional cloud architectures are hitting their limits. Meeting this moment requires purpose-built infrastructure: AI clouds  designed for performance, scale, and security from the ground up. At CoreWeave, we’ve been building exactly that from the beginning: the Essential Cloud for AI. Every architectural decision we make is about giving innovators the fastest, most efficient path from idea to production. That’s why we’ve invested deeply in hardware acceleration, and why the next-generation NVIDIA BlueField-4 DPU is a natural evolution for our platform… and catalyst for your next big breakthrough. NVIDIA BlueField-4: CoreWeave’s next step forward in building the #1 AI Cloud NVIDIA BlueField-4 delivers a generational jump in DPU performance, with 6x the compute power and 800 Gb/s of throughput. But what excites us most isn’t just the raw specs,  it’s how this game changing technology amplifies what CoreWeave already does best: building high-performance, secure, and elastic infrastructure for AI workloads at scale. We’ve been integrating BlueField DPUs into every CoreWeave server for years, offloading networking, security, and storage management tasks from CPUs so every GPU can stay focused on what matters — training and inference. BlueField-4 pushes that efficiency even further, unlocking faster data access, lower latency, and stronger tenant isolation in multi-tenant AI clusters. Engineering differentiation: How CoreWeave Cloud uses BlueField Our DPU strategy is a fundamental part of our architecture, not an afterthought. Working closely with NVIDIA, we’ve used the NVIDIA DOCA software framework to push core networking and security functions directly onto BlueField DPUs. This lets us build hardware-enforced tenant isolation , using EVPN VXLAN to create separate, secure virtual networks for each customer. The result: no software bottlenecks, no noisy neighbors, and no compromises. It’s the difference between a general-purpose, legacy hyperscaler cloud and one specifically  purpose-built for AI.. What NVIDIA BlueField-4 on CoreWeave unlocks for customers Extreme performance and scale: BlueField-4 gives 6x performance boost and 800 Gb/s throughput will supercharge CoreWeave’s networking and storage fabrics, delivering faster model access, lower inference latency, and the ability to handle the most demanding AI workloads.

Next-gen security: With the new BlueField Advanced Secure Trusted ResourceArchitecture we’ll extend our zero-trust infrastructure model with even finer-grained, hardware-based isolation.

Maximum efficiency: Every workload cycle matters. By shifting even more infrastructure tasks to the DPU, BlueField-4 ensures more CPU and GPU cycles are dedicated to our customers’ compute jobs, driving higher utilization and better economics.

Building the Essential Cloud for AI together BlueField-4 validates the approach we’ve been perfecting for years: infrastructure that fades into the background, allowing developers, researchers, and enterprises to move faster than ever before. At CoreWeave, we’re not just adopting the latest technology,  we’re architecting the CoreWeave Cloud around it. By deeply integrating NVIDIA BlueField-4 into our platform, we’re setting a new standard for how modern AI clouds should perform: faster, more secure, and more efficient at scale. This is how CoreWeave Cloud powers the future of AI — built-for-purpose, performance-obsessed, and engineered for the pioneers  defining what’s next. Learn more about why CoreWeave is your best choice for running complex AI workloads. Find out why CoreWeave is the only AI Cloud leading MLPerf results in both training and inference. Talk to one of our AI experts to learn how CoreWeave can get your AI breakthroughs to market faster. ‍

CoreWeave is advancing AI cloud performance with NVIDIA BlueField-4, delivering faster, more secure, and efficient infrastructure purpose-built for next-gen AI innovation.

Share this article: Copied

Related Blogs

The Data Center Questions Everyone Is Asking, Answered 5 min read

What a Reference Architecture for Distributed AI Training Actually Looks Like 6 min read

Why Inference Latency and Availability Drift in Production 7 min read

5 Misunderstandings About Enterprise AI Training Infrastructure 5 min read

Choosing the Right NVIDIA Platform for Running Inference on CoreWeave 5 min read

CoreWeave Closes the Loop Between Training and Inference 4 min read

Why Distributed Training Fails at Scale 7 min read

Run Agentic Workloads Safely at Scale with CoreWeave Sandboxes 6 min read

Red Hat AI Inference on CKS for Hybrid Inference 4 min read

CoreWeave Is Now the Fastest at Inference on One of the Best Open Source Models Kimi K2.6 3 min read

Contact us Login

Products GPU Compute CPU Compute Storage Services Networking Services Managed Services Bare Metal Servers Platform Fleet LifeCycle Controller

Node LifeCycle Controller Tensorizer Observability

Solutions AI Model Training AI Inference VFX & Rendering Mission Control

AI Infrastructure

Why CoreWeave

Resources Customer Stories Documentation Status Pricing Resource Center Events & Webinars

About About Us Careers Life at CoreWeave

Newsroom Investor Relations Supplier Code of Conduct Terms of Service Do Not Sell or Share My Personal Information

© CoreWeave. 290 W Mt Pleasant Ave Suite 4100 Livingston, NJ 07039

CoreWeave Cloud,

Copy code Copied!