Skip to main content
Sam Altman, OpenAI CEO, stands in a data center aisle, servers glowing, symbolizing the AI RAM supply squeeze.

Editorial illustration for Sam Altman: data centers better than desks as OpenAI RAM spree squeezes supply

OpenAI's DRAM Spree Squeezes Memory Chip Supply

Sam Altman: data centers better than desks as OpenAI RAM spree squeezes supply

• 4 min read

Memory chips are getting scarce, and Sam Altman has a theory about who's hoarding them wrong. OpenAI has spent the past year buying up DRAM at a pace that's rattled the rest of the tech industry, with executives at Micron and SK Hynix pointing to the company's data center build-out as a reason chip prices keep climbing. That scarcity is starting to show up everywhere, from gaming rigs to the RAM sitting inside your laptop.

Altman addressed the shortage directly during an appearance on the Cheeky Pint podcast, where he also weighed in on how AI is changing math. The question put to him was straightforward: why does OpenAI need so much RAM, and what does that mean for everyone else competing for the same limited supply? His answer cuts against a common assumption in tech circles, the idea that open-source models running locally on personal hardware will ease the pressure on data centers. Altman argued the opposite is true, and that the logic people use to justify keeping memory on their own machines gets the whole allocation problem backwards.

Right now, memory companies Samsung and SK Hynix have long term deals with cloud service operators and projects like OpenAI’s Stargate that have reserved much of the world’s RAM output over the next few years. And the billions they’re willing to lay out for chips is why the laptops, phones, and other tech you use are getting more expensive right now.

Why this matters

Altman is making a bet that every reader building on AI infrastructure should scrutinize: that centralized compute always beats distributed compute, even as his own company's purchasing is what's driving the RAM shortage in the first place. That's a convenient argument for a CEO whose business model depends on everyone running inference through his data centers rather than their own machines. The open-source crowd he's dismissing has a real point, local models on local hardware don't compete with OpenAI for the same memory chips, and they don't require anyone to trust a third party with their data or their compute budget.

For developers and founders, this is worth watching closely. If RAM scarcity becomes the new normal because a handful of labs are hoarding it for training runs, that's a tax on everyone else building smaller, local, or open systems. Altman's framing treats that scarcity as a feature, not a cost he's imposing on the rest of the market.

Common Questions Answered

Why is OpenAI's massive DRAM purchasing causing memory chip shortages across the tech industry?

OpenAI has been aggressively buying up DRAM chips at unprecedented rates to support its data center expansion and infrastructure projects like Stargate, which has reserved much of the world's RAM output over the next several years. This massive purchasing power has rattled the rest of the tech industry and driven up memory chip prices, making components more expensive for consumer devices like laptops, gaming rigs, and phones.

What is Sam Altman's argument about centralized versus distributed compute in AI infrastructure?

Altman argues that centralized compute in data centers is superior to distributed compute on local machines, suggesting that AI infrastructure should be concentrated rather than decentralized. However, critics point out this argument is convenient for OpenAI's business model, which depends on users running inference through the company's data centers rather than on their own hardware.

Which memory chip manufacturers have long-term deals with cloud operators and AI projects like OpenAI's Stargate?

Samsung and SK Hynix have secured long-term deals with cloud service operators and major AI projects including OpenAI's Stargate, which have reserved significant portions of the world's RAM output over the coming years. These multi-billion dollar commitments from tech companies are a primary driver of the current memory chip scarcity affecting consumer electronics.

How is the RAM shortage affecting consumer technology prices?

The billions that memory companies are committing to supply cloud operators and AI infrastructure projects like OpenAI's have reduced the available RAM supply for consumer devices, directly driving up prices for laptops, phones, and other consumer tech. This shortage is visible across the entire consumer electronics market as manufacturers compete for limited memory chip inventory.

LIVE22:54Microsoft records USD 18.6B free cash flow as AI spending boom cracks emerge