Live data from Hacker News

The RAM shortage could last years

theverge.com

1–10 of 536 posts

Re: The RAM shortage could last years

#3
I'm a bit surprised the article makes no mention of Google's TurboQuant[0] introduced 26 days prior.

Given that TurboQuant results in a 6x reduction in memory usage for KV caches and up to 8x boost in speed, this optimization is already showing up in llama.cpp, enabling significantly bigger contexts without having to run a smaller model to fit it all in memory.

Some people thought it might significantly improve the RAM situation, though I remain a bit skeptical - the demand is probably still larger than the reduction turboquant brings.

[0] https://news.ycombinator.com/item?id=47513475

Re: The RAM shortage could last years

#7
post #3

I'm a bit surprised the article makes no mention of Google's TurboQuant[0] introduced 26 days prior. Given that TurboQuant results in a 6x reduction in memory usage for KV caches and up to 8x boost in speed, this optimization is already showing up in llama.cpp, enabling significantly bigger contexts without having to run a smaller model to fit it all in memory. Some people thought it might significantly improve the R…

You can still use as much memory, but fit more things into it, so I don’t think the current market hogs will let go easily.

Re: The RAM shortage could last years

#8
post #3

I'm a bit surprised the article makes no mention of Google's TurboQuant[0] introduced 26 days prior. Given that TurboQuant results in a 6x reduction in memory usage for KV caches and up to 8x boost in speed, this optimization is already showing up in llama.cpp, enabling significantly bigger contexts without having to run a smaller model to fit it all in memory. Some people thought it might significantly improve the R…

You still need to hold the model in memory. If you have for example 16 GB ram, the gains aren't that much
Post reply on HN