Archive — US — techcrunch
OpenAI CEO Sam Altman says the company is ‘out of GPUs’
OpenAI CEO Sam Altman announced the company is 'out of GPUs,' delaying the rollout of GPT-4.5. The new model, requiring tens of thousands more GPUs, will initially be available to ChatGPT Pro subscribers. OpenAI faces significant computing capacity challenges and aims to resolve this by creating its own AI chips and expanding its data center infrastructure.
Key points
- Sam Altman stated OpenAI is currently out of GPUs, impacting GPT-4.5 rollout.
- GPT-4.5 will first be available to ChatGPT Pro subscribers, followed by ChatGPT Plus users.
- The model's operational costs are high; OpenAI's pricing for GPT-4.5 is significantly increased compared to GPT-4.
- Altman cited rapid growth as a reason for GPU shortages, which they did not predict.
- OpenAI plans to tackle future GPU shortages by developing proprietary AI chips and expanding their data center network.
Why it matters
- In 2025, there is a trend of increasing demand for powerful GPUs due to advancements in AI models.
- Many tech firms are currently facing similar supply chain challenges related to hardware for AI.
- The transition towards in-house hardware production reflects a growing trend in the industry to mitigate reliance on third-party suppliers.
OpenAI's current GPU shortage is a significant setback, but their plans for AI chip development and data center expansion may bolster future growth, ensuring they remain competitive in the fast-evolving AI landscape.