Nikhil Cherian, Senior Director of Supply Chain Infrastructure at Google, recently pointed out at a summit forum that DRAM supply is extremely tight. Google has already begun recycling DDR4 memory from retired servers and integrating it into the latest-generation AI servers. He further admitted that even this measure is still not enough, and called on memory manufacturers to rapidly expand production capacity.
Nikhil Cherian stated bluntly that for cloud service providers, if AI processors built at a cost of billions of dollars sit idle while waiting for memory bandwidth, it is equivalent to wasting a massive amount of compute capital expenditure. Therefore, Google's current priority is not simply memory pricing, but rather whether there is sufficient memory supply to keep expensive AI chips running at full capacity on a sustained basis.
