What's rapidly changing are quantization algorithms, and hardware features to support those algorithms. For example, Blackwell GPUs support dynamic FP4 quantization with group size 16. At that group size it's close to lossless (in terms of accuracy metrics).
replyBoth AMD and Nvidia are dumping more and more memory into their GPUs.
replyMI300x is 192GB HMB3, MI325x is 256 HMB3e, MI355x should be 288 HBM3e (and support FP4/6).
The professional side of things, yes. For consumer grade GPUs, despite the trends in gaming markets otherwise needing such, the values have stagnated a bit.
replyI'm NDA with AMD and sadly can't mention details, but I can say the future is promising.
replyMusic to my ears. The entire market needs more competitors. As a happy Ryzen owner, I look forward to it.
replyAs long as AMD fixes the damn driver issues I've seen for over a decade.
I hope AMD cracks the CUDA Problem soon
replyI'm personally really excited about this solution: https://docs.scale-lang.com/
replyYes, yes.
replyNvidia about to release blackwell ultra with 288GB. Go back to maybe 2018 and max was 16gb if memory serves.
DeepSeek recently release a 670 gb model. A couple years ago Falcon's 180gb seemed huge.
I'd assume that, in the context of LLM inference, "recent" generally refers to the Ampere generation and later of GPUs, when the demand for on board memory went through the roof (as, the first truly usable LLMs were trained on A100s).
replyWe've been stuck with the same general caps on standard GPU memory since then though. Perhaps limited in part because of the generational upgrades happening in the bandwidth of the memory, rather than the capacity.
Bandwidth is going up too. "It's not doubling every 18 months and hence it's not moving" isn't a sensible way to view change.
replyA one time effective 30% reduction in model size simply isn't going to be some massive unlocker, in theory or in practice.