Not sure I follow the logic. Sounds like you don’t think additional supply would result in lower prices, all else equal.
But it will of course be better than if no supply is added. Microsoft and Google would want the RAM you mentioned regardless, so if no supply is added the situation will be even worse, because then they’ll purchase it from somewhere else driving price higher.
It isn't that I don't think that it would make a difference.
Microsoft and Google would undoubtedly slurp up any excess capacity initially (I'd be shocked if they didn't tie in as much production as possible for 12-18 months, just like OpenAI did).
Thus, unless it was a huge supply increase, it might be a year or two after that before consumer DRAM supplies begin to increase. And that is presuming we are talking about DDR5 - if all this happens at the time of DDR6 introduction, all bets are off as the AI companies will want to shift over as they upgrade their data centers.
What I am saying is that any new memory supplier entering the market now would be risking a couple billion dollars while not seeing any possibility of a return on investment until 3-6 years from now.
And, given the peculiar nature of the AI craze (not to mention the past historical cyclical boom/bust nature of the DRAM memory market), I just don't see anyone crazy enough in the current economy to risk jumping on that "invest in a new memory fab" bandwagon anytime soon.
Honestly, the best thing that could happen for us is for OpenAI to loose a lot of money, fire Altman, and have to sell everything they stockpiled that they never needed in the first place.
Well... DDR6 shift might get delayed and focused on servers, client side would wait for more time.
The problem with that is that the memory manufacturers won't just build new fabs for DDR6 -- it is far more cost effective for them to shift their existing fab lines currently producing DD5 over to DDR6. There will be a period of simultaneous DDR5/DDR6 production as they shift (just like there was with DDR4/DDR5), but the fabs will try to quickly shift to DDR6 as soon as DDR6 yields increase to acceptable levels because the manufacturers will be able to charge a lot more money for faster DDR6.
DDR5 memory production will then will then taper off and die, just like DDR4 production did after DDR5 yields were finally maximized.