OpenAI Launches Guaranteed Capacity for AI Compute Access
Quick Look
- OpenAI introduced Guaranteed Capacity, a new offering allowing customers to secure long-term access to AI compute power.
- Available in one, two, or three-year commitments with increasing discounts, the service aims to provide capacity certainty amid growing AI demand.
AI-generated summary
Why It Matters
OpenAI is introducing a new offering called Guaranteed Capacity to address the increasing demand for AI compute power. This move comes as the AI industry faces significant capacity constraints and OpenAI prepares for a potential IPO, having previously secured multi-billion dollar compute deals.
OpenAI on Tuesday announced a new offering called Guaranteed Capacity, which allows customers to secure long-term access to compute to power their artificial intelligence products, agents and workflows.
The company said customers can choose between one, two and three-year-long commitments, and it's offering discounts that increase based on those annual commitments, according to its website.
"Customers are increasingly asking us for certainty on capacity. As models get better, we expect that the world will be capacity-constrained for some time," OpenAI CEO Sam Altman wrote in a post on X.
Altman said the new offering will help OpenAI plan ahead, which he hopes will be a "big win-win." He added that OpenAI will offer Guaranteed Capacity until it sells out of its current allocation, but that the company plans to offer it again in the future.
In the AI industry, compute refers to the computational power and resources required to train and run large AI models. It's extremely expensive and difficult to build on a large scale. OpenAI has told investors that it is targeting roughly $600 billion in total compute spend by 2030, as CNBC previously reported.
OpenAI, which is is valued at more than $850 billion by private investors, is looking to bring in more revenue as it gears up for a potentially massive IPO as soon as this year. The company unnerved Wall Street by inking a flurry of multi-billion-dollar compute deals late last year, sparking debates about how OpenAI would be able pay for such a large infrastructure buildout.
Altman repeatedly brushed off concerns, writing in a post on X in November that OpenAI expects to grow to hundreds of billions in sales by 2030. The company's new Guaranteed Capacity program could help lay the foundation for what its compute business model will eventually look like.
In his post on Tuesday, Altman said OpenAI will make sure it leaves enough capacity available for its products like ChatGPT and its coding assistant Codex.
Open Questions
- What is the exact capacity allocation for Guaranteed Capacity?
- What are the specific pricing tiers and discount structures?
- How will OpenAI balance Guaranteed Capacity with its own product needs like ChatGPT?
- What are the long-term implications for smaller AI developers unable to afford such commitments?
