Recent pronouncements from Elon Musk suggest a future where robot RAM requirements could outstrip all other demands on Earth. However, a closer examination of these claims indicates that the current and projected needs of AI data centers will continue to far exceed those of robots for the foreseeable future, potentially mitigating the impact on the ongoing global memory shortage.
The discussion was sparked by Musk's assertion that "Robot RAM needs will exceed everything else on Earth, as you need the robots to be on physical substates in order to shape atoms." He further noted that "Super intelligence needs will vastly exceed that, but ~100% of SI will be deep space compute. Obviously." This statement, made in response to a report predicting significant DRAM growth for humanoid robots, has raised concerns among PC hardware enthusiasts already grappling with inflated prices due to AI-driven demand.
The Counterpoint Research report, referenced by Musk, reportedly claims that by 2030, humanoid robots could demand 20 times more DRAM than they do currently. This would translate to an increase from 8,300 terabytes (Tb) to 172,000 Tb. While this represents a substantial jump in robot-specific memory consumption, it remains a small fraction of the global DRAM capacity.
Global DRAM capacity is currently estimated to be around 40 exabytes (EB), which equates to 320,000,000 terabits (Tb). The projected 172,000 Tb for humanoid robots by 2030 would constitute only about 0.05% of the current global DRAM capacity. Even with anticipated growth in overall DRAM production and consumption by 2030, this figure is expected to remain a minor portion compared to the demands of AI.
Previous analyses and statements from memory manufacturers like SK Hynix have already predicted a memory supply shortfall extending beyond 2030 due to persistent AI demand. The added demand from humanoid robots, while notable, is unlikely to significantly alter the market dynamics. For robot-related RAM needs to have a substantial impact on the current shortage, it would likely necessitate a scenario where AI data centers are relocated, such as to space as Musk alluded to, which is a distant prospect.
Therefore, the primary driver of the current memory shortage and its impact on PC component prices remains the insatiable demand from the AI sector. The burgeoning field of robotics, while growing, does not yet pose a comparable threat to global memory supply or pricing.