Humanoid robot DRAM demand forecast to rise 20-fold by 2030

By Splitfeed EditorialLast updated Oct 16, 2026, 08:30 UTC

Around October 15 and 16, 2026, forecasts circulated projecting that DRAM demand generated by humanoid robots will grow roughly 20 times its current level by 2030. Samsung and SK Hynix were identified as the memory manufacturers best placed to capture that growth. The projection was reported by Digitimes and TokenPost, though the underlying forecast source, the methodology behind the estimate, and the baseline unit volumes were not specified in the available reporting.

The significance of this forecast lies in what it implies about humanoid robots as a category of memory consumer. Data centers and consumer electronics have historically driven DRAM demand cycles, and memory makers have geared their capital expenditure and product roadmaps around those customers. If humanoid robots do emerge as a high-volume, memory-intensive application at the scale suggested, it would represent a structurally new demand curve, one that could reshape procurement strategies and capacity planning across the semiconductor supply chain. What remains genuinely in dispute is whether humanoid robot deployment will reach a volume and pace by 2030 that makes a 20-fold increase realistic rather than aspirational.

None of the three camps have published reactions to this story yet. The Pro-AI camp would typically treat a forecast like this as evidence that physical AI systems are maturing into a serious economic force, vindicating continued investment in robotics and the supporting hardware ecosystem. The Anti-AI camp would likely raise skepticism about the deployment timeline, question whether humanoid robots will reach mass adoption by 2030, and flag the risk that such forecasts serve the commercial interests of the memory companies named as beneficiaries. The Middle Ground camp would generally call for scrutiny of the methodology behind the 20-fold figure and caution against treating a single demand projection as a settled roadmap.

The figures that would most sharpen or soften this forecast are humanoid robot shipment volumes from major manufacturers over the next one to two years, along with any revised guidance from Samsung or SK Hynix that explicitly ties capacity investment to robotics demand rather than data-center AI alone.