Backblaze and CoreWeave Expand AI Infrastructure with $335M Multi-Exabyte Storage Partnership

Backblaze has announced a 5-year, multi-exabyte data storage agreement with CoreWeave valued at approximately $335M. Under the partnership, Backblaze will provide storage capacity supporting portions of CoreWeave's managed storage infrastructure, expanding the storage architecture behind AI workloads and enterprise-scale data management.

Backblaze is an independent cloud storage company founded in 2007 and headquartered in San Mateo, California. CoreWeave is an AI cloud infrastructure provider serving leading AI developers and enterprise customers. Together, the companies are extending a relationship that began years ago into a commercial agreement designed to support increasingly data-intensive AI environments.

The announcement matters because AI infrastructure is entering a new phase. The industry conversation has focused almost exclusively on GPUs and compute capacity, yet storage has quietly become one of the defining constraints on AI performance, operating costs, and scalability. This partnership demonstrates that storage architecture is becoming as strategically important as compute itself.

About Backblaze

Infrastructure rarely becomes the headline. It becomes the headline only when it disappears. Backblaze has spent years building exactly the kind of business most people ignore until everything stops working. GPUs capture attention. Foundation models dominate conference stages. Storage simply makes the entire machine function.

Founded in 2007, Backblaze started with 7 founders working from a Palo Alto apartment and has grown into a publicly traded cloud storage company serving more than 500,000 customers across 175 countries. Its Backblaze B2 Cloud Storage platform alone serves more than 100,000 customers worldwide. Rather than competing to build another AI model, Backblaze has focused on becoming the infrastructure layer beneath them, with a mission centered on helping customers solve complex data storage challenges.

That positioning has become increasingly relevant as AI workloads produce massive volumes of training data, checkpoints, model outputs, retrieval datasets, and long-term archives. Compute may generate the headlines, but storage determines whether those workloads remain economically sustainable.

What Happened

The new agreement significantly expands the relationship between Backblaze and CoreWeave. Backblaze will provide cost-efficient, HDD-based storage capacity supporting portions of CoreWeave's managed storage infrastructure. The technology supports storage tiers within CoreWeave AI Object Storage, allowing customers to access additional storage tiers without modifying application code.

On the surface, this looks like another enterprise infrastructure partnership. It is considerably more important than that.

Modern AI environments constantly balance performance against economics. Flash storage delivers exceptional speed but comes at a premium, while HDD-based storage dramatically lowers costs for large datasets that do not require constant high-performance access. Intelligent storage tiering allows organizations to keep premium storage available for active workloads while moving colder data onto lower-cost infrastructure. That balancing act is becoming a defining element of AI infrastructure strategy.

Leadership Signals

Leadership commentary from both organizations reinforces the broader strategic direction behind the agreement.

Gleb Budman, Co-Founder, CEO, and Chairperson of Backblaze, emphasized that storage forms the foundation beneath every AI workflow, while Nick Hoover, Vice President at CoreWeave, highlighted Backblaze's ability to simplify large-scale HDD-based infrastructure while maintaining enterprise reliability. Marc Suidan, CFO at Backblaze, also pointed to continued momentum surrounding the company's neocloud storage offering as enterprise AI demand accelerates.

Taken together, these perspectives reflect a broader market reality. Infrastructure vendors increasingly recognize that AI competitiveness depends on complete system design rather than isolated hardware performance.

Why This Matters

For much of the AI boom, investors rewarded whoever could deploy the most GPUs. That narrative is beginning to mature. Training large language models requires enormous amounts of data movement. Inference generates continuous storage demands. Enterprise deployments introduce governance requirements, archival needs, retrieval systems, and compliance obligations that extend well beyond compute.

Storage has quietly evolved from an operational necessity into strategic infrastructure. The Backblaze-CoreWeave partnership illustrates that transition. Instead of asking only who owns the fastest accelerators, enterprise buyers are increasingly evaluating how efficiently AI platforms manage data throughout its lifecycle. Storage architecture now directly influences infrastructure costs, application responsiveness, scalability, and long-term profitability.

Infrastructure is becoming an integrated system rather than a collection of individual technologies.

Market Context

The AI infrastructure market continues to diversify. Cloud providers, storage companies, networking vendors, and GPU operators are increasingly collaborating to deliver complete AI environments instead of isolated services. Organizations deploying AI at enterprise scale need reliable movement between storage, networking, compute, inference, and long-term retention, creating opportunities for companies like Backblaze that specialize in one critical layer of the stack instead of attempting to build everything themselves.

The agreement also reflects another industry shift. Enterprise customers increasingly prefer interoperable infrastructure that can evolve without forcing major application rewrites. Existing CoreWeave AI Object Storage customers can access additional storage tiers without changing application code, reducing operational friction while expanding deployment flexibility.

That kind of operational simplicity often creates competitive advantages that benchmark charts never capture.

What This Signals

Every major technology cycle eventually reaches the same point. The obvious winners attract attention first. The indispensable infrastructure providers emerge later.

Backblaze is positioning itself deeper inside AI infrastructure rather than remaining solely a cloud storage provider, while CoreWeave continues expanding beyond GPU access into a more comprehensive AI cloud platform. Together, the companies illustrate how the AI stack is becoming increasingly modular, specialized, and interconnected.

The biggest lesson may not be the $335M headline. It may be that storage has stopped being background infrastructure. It is now part of the competitive strategy.