So you're arguing that ai companies overbought memory to increase scarcity? Possibly even buying e.g. consumer ddr?
happyPersonR 2 hours ago [-]
Hyperscalers and social media companies hired tech talent to keep folks from starting their own companies and keep other companies from hiring them.
memory companies are known to price fix as well, so i could really see either possibility tbh
and why not both?
theoneone 5 hours ago [-]
just a theory! The end result was too make it difficult for me and you to run llms on our systems, and you need porwerfull system to run them. But some months ago, those powerfull systems where affordable for end consumers. Now, good luck buying let's say 128gb ddr5.
MildlySerious 3 hours ago [-]
Do you believe this is more likely than corporations trying to restrict the supply for their corporate competitors? My thinking would be that we're just caught in the crossfire. The percentage of people seriously running their own models locally would hardly be any less diminishing if hardware prices were at their previous levels.
theoneone 3 hours ago [-]
absolutely!
ukblewis 2 hours ago [-]
You are aware that a 1 trillion parameter model won’t run on any quantity of consumer hardware, even heavily quantized… you need a data center for that. And that is not even to talk of the Kimi K3 model which is in the ballpark of 3 trillion
layla5alive 2 hours ago [-]
You do not need a data center - just a single rack will do to run it fast. Or about 100x 5090s.
theoneone 11 hours ago [-]
Some personal thoughts on the current ram/GPU shortage -price hikes.
memory companies are known to price fix as well, so i could really see either possibility tbh
and why not both?