CoreWeave Secures Priority Solidigm SSD Capacity
On August 5, CoreWeave signed a multi-year agreement giving it priority access to Solidigm enterprise SSD capacity, supporting storage growth alongside demand for its AI cloud platform. Financial terms were not disclosed, SDxCentral reported, while industry coverage cited tightening NAND supply and rising contract prices. The agreement adds storage procurement to the infrastructure planning considerations surrounding large-scale AI cloud operations.
CoreWeave signed a multi-year agreement with Solidigm on August 5 for priority access to enterprise solid-state drive capacity. According to CoreWeave's announcement, the arrangement is intended to secure dedicated storage supply as demand grows for the company's AI cloud platform; neither company disclosed financial terms.
The agreement covers enterprise SSD allocation rather than a specified volume of drives, spend commitment, pricing mechanism, or delivery schedule. CoreWeave described storage as a capacity-planning constraint amid accelerating enterprise AI adoption and tightening industry storage supply.
"This agreement gives us priority access to the storage our roadmap depends on, allowing customers to focus on what they're building instead of the infrastructure underneath it," CoreWeave COO Sachin Jain said in the company release.
Solidigm executive vice president and head of global sales Paul Palonsky said long-term supply commitments are becoming a strategic advantage and that the deal reflects Solidigm's confidence in CoreWeave's growth. SDxCentral also reported that the agreement gives CoreWeave priority access to Solidigm's enterprise SSD capacity, while noting that commercial terms were not made public.
Storage joins compute in AI capacity planning
The announcement makes SSD supply a more explicit part of AI-cloud procurement. GPU availability has been the most visible infrastructure constraint for model training and inference, but large AI environments also require substantial storage for datasets, checkpoints, model artifacts, caching, and high-throughput data pipelines.
StorageReview characterized the deal as a response to flash allocation becoming an AI-cloud planning issue. Citing TrendForce forecasts, the publication reported projected NAND flash contract-price increases of 10-15% quarter over quarter in the third quarter of 2026, alongside a forecast of a NAND shortage through 2026. StorageReview further reported that meaningful new NAND capacity is unlikely before late 2027 or 2028.
Those market forecasts are not terms of the CoreWeave-Solidigm agreement. They nevertheless provide context for why a buyer may value allocation certainty even when enterprise SSD supply conditions differ by product category. StorageReview noted that enterprise SSD availability has improved as suppliers shift capacity away from consumer segments, while its cited TrendForce outlook still described broader NAND supply as constrained.
What remains undisclosed
The public disclosures leave several operational details open, including:
- •The amount of SSD capacity reserved for CoreWeave.
- •The NAND technologies, drive form factors, and performance specifications involved.
- •Whether the agreement includes fixed prices, minimum purchase commitments, or volume flexibility.
- •The extent to which the allocation supports training storage, inference caching, managed storage, or other workloads.
SDxCentral noted that CoreWeave has previously expanded its storage ecosystem through a Pure Storage agreement for managed storage infrastructure, a partnership with Vast Data, and its Local Object Transport Accelerator, or LOTA, storage technology.
For ML infrastructure teams, the deal illustrates a broader procurement pattern: in constrained hardware markets, compute reservations alone may not guarantee deployable capacity. Comparable AI-cloud buildouts increasingly depend on aligning accelerator availability with network, memory, and storage throughput so that data movement and checkpoint operations do not become the limiting layer.
Key Points
- 1CoreWeave obtained priority access to Solidigm enterprise SSD capacity, extending its AI infrastructure procurement beyond accelerator supply.
- 2Commercial terms and reserved volumes remain undisclosed, limiting visibility into the agreement's direct financial and capacity impact.
- 3Comparable AI-cloud deployments increasingly require coordinated compute, networking, memory, and storage procurement to avoid infrastructure bottlenecks.
Scoring Rationale
The agreement is a notable AI infrastructure supply arrangement involving a major GPU cloud provider and enterprise SSD vendor. It reinforces storage allocation as an operational consideration for large-scale AI platforms, although undisclosed volumes and economics limit its immediate measurable impact for practitioners.
Sources
Primary source and supporting public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
