OpenAI CEO Sam Altman has argued that the world's limited supply of DRAM could be used more efficiently in data centres, even as rising memory prices squeeze PC gamers and smartphone buyers.
OpenAI · Memory supply debate
Sam Altman says data centres could use scarce DRAM more efficiently
His argument targets a squeeze that reaches beyond the cloud: the same memory that powers AI clusters also drives the cost of PCs and smartphones.
How the pressure travels
A single component pool feeds both sides of the market, so demand growth in one segment reaches the other.
AI clusters
Training and inference consume large volumes of DRAM.
→
Tighter supply
Competition for the same chips limits what others can buy.
→
Costlier devices
Gamers building PCs and smartphone buyers face higher memory costs.
The case for efficiency
Data centres already consume DRAM; using it better could ease demand.
Extracting more computing work per gigabyte reduces pressure on consumers.
Efficiency gains would mean fewer buyers competing for a fixed supply.
Open questions
The specific engineering or purchasing measures have not been detailed.
It is unclear how much of the gain is achievable at scale.
No timeline has been given for any effect on consumer memory prices.
What would have to change
Efficiency only eases consumer prices if data centre practice shifts in concrete ways.
Software optimisation
Code that extracts more work from existing memory.
Workload design
How AI jobs are structured and scheduled.
Hardware choices
Build decisions that take time to implement.
Watch for whether these changes arrive, and whether consumer memory prices ease in the coming quarters.
Continue reading The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in ✓ Signed in — this article isn’t included in your current plan.Unlocking the full article...