ChainCatcher reports that multiple Silicon Valley venture capital firms are racing to secure GPU computing power to provide dedicated resources for their AI startups. a16z established a GPU compute pool of approximately 20,000 GPUs as early as 2024 and is currently negotiating expansion with data center companies Switch and Volta. General Catalyst plans to build a dedicated data center for its portfolio companies and hired the former head of NVIDIA’s DGX Cloud in May to oversee its AI infrastructure operations. Even after securing funding, AI startups may still struggle to rent GPUs in a timely manner. In April this year, it was revealed that major cloud providers like Microsoft prioritized computing resources for their internal teams and large clients, leaving some startups waiting for months. a16z previously disclosed that during periods of compute scarcity, prices originally available under six-month contracts may require three-year commitments. Model training demands short-term, intensive GPU usage, while inference needs are difficult to predict—locking in multi-year compute capacity risks capital inefficiency. Through its Oxygen project, a16z centrally procures and leases computing power, then allocates it to its portfolio startups, granting them pricing and contract terms typically reserved for enterprise clients. In August this year, a16z co-led a $300 million funding round for Volta, which reached a $2.4 billion valuation; the investment simultaneously supports both AI startups and the data center providers supplying compute power. Small VC Crucible Capital has built a cluster featuring 576 NVIDIA B300 GPUs for its portfolio startups to test on, while also selling excess capacity to a cutting-edge AI lab. This model shifts the financial burden previously borne primarily by cloud providers and data center operators onto venture capital firms.
Silicon Valley VCs Secure GPU Capacity for AI Startups
ChaincatcherShare
Summary
Technical analysis for crypto investors is increasingly centered on the hardware race, as Silicon Valley VCs secure GPU capacity for AI startups. a16z has assembled a GPU pool of 20,000 cards and is expanding with Switch and Volta. General Catalyst is building a data center and has hired NVIDIA’s former head of DGX Cloud. Despite funding, startups still encounter GPU shortages, as cloud providers prioritize large clients. a16z now requires three-year contracts during periods of scarcity; its Oxygen project offers improved pricing. Crucible Capital operates a 576-GPU cluster and sells excess capacity. The risk-to-reward ratio for early-stage AI projects remains high amid constrained hardware availability.
Source:Show original
Disclaimer: The information on this page may have been obtained from third parties and does not necessarily reflect the views or opinions of KuCoin. This content is provided for general informational purposes only, without any representation or warranty of any kind, nor shall it be construed as financial or investment advice. KuCoin shall not be liable for any errors or omissions, or for any outcomes resulting from the use of this information.
Investments in digital assets can be risky. Please carefully evaluate the risks of a product and your risk tolerance based on your own financial circumstances. For more information, please refer to our Terms of Use and Risk Disclosure.