TECH NEWS – The company best known for ChatGPT could gain an enormous lead over its rivals if it actually pulls this off.

Amazon and Anthropic have announced plans to bring 6 GW of AI compute capacity online by the end of the year. Their main rival, OpenAI, is thinking on a much larger scale. Sam Altman’s company says that since the arrival of ChatGPT, both revenue and compute resources have been growing rapidly. That growth has been driven entirely by demand for AI, and as the industry moves into its next phase, service providers are seeing even more interest. To satisfy that enormous demand and the huge wave of users entering the AI ecosystem, OpenAI is preparing to expand its compute capacity by nearly 16 times by 2030 compared with its 2025 level. In 2025, the company was operating at roughly 1.9 GW of AI compute. Back in January of last year, it committed to building out 10 GW, and it has already identified more than 8 GW of that figure.

In January 2025, we committed to generating 10GW of compute and have already identified over 8GW of that. Now, we’re planning for 30GW of compute by 2030. A milestone that scales with the rapidly accelerating demand for intelligent systems.

Image generated by @ChatGPTapp Images… pic.twitter.com/h6UMgG5UCV

— OpenAI Newsroom (@OpenAINewsroom) April 22, 2026

OpenAI is now aiming for 30 GW of compute power by 2030. That would represent a massive boom for the semiconductor industry, because rising demand on that scale would inevitably drive the construction of new factories, new power facilities, and large numbers of new jobs around the world. A recent OpenAI patent has also surfaced, outlining plans for a custom AI chip featuring multiple HBM stacks, apparently far more than what has been seen in current products on the market. Those chips could very well become part of the broader 30 GW expansion plan.

The problem is that beyond the AI silicon itself, the core ingredient in those chips is HBM, and that is already in very short supply. That is why Samsung and SK Hynix are racing to build more fabs and expand capacity. On paper this sounds great, but it is also impossible to ignore the strain current AI demand is placing on the wider tech sector. As AI keeps growing, consumers are increasingly likely to feel it through rising prices on all kinds of products, including smartphones, PCs, and game consoles.

The semiconductor manufacturing sector therefore has to undergo a major transformation if it wants to create smarter and more efficient production capacity capable of satisfying not just today’s needs, but tomorrow’s as well.

Source: WCCFTech