Stability AI Launches the First of Its StableLM Suite of Language Models

Re:Stability AI Launches the First of Its StableLM Suite of ... by brucethemoose2 on Wednesday April 19, 2023 @06:36PM (#35634185)

The M2 and the 4090 are both very general purpose. In fact, the 4090 allocates proportionally more silicon area to the tensor cores than Apple allocates to the neural engine.

The M series is basically the only "big" SoC with a functional, flexible NPU and big GPU right now, which is why it seems so good at ML. But you can bet actual ML focused designs are in the pipe.

[ Parent | on HN ]

Re:Stability AI Launches the First of Its StableLM Suite of ... by zamnos on Wednesday April 19, 2023 @06:48PM (#35634278)

Well, and gobs of RAM, which only top end cards can compete with.

[ Parent | on HN ]

Re:Stability AI Launches the First of Its StableLM Suite of ... by brucethemoose2 on Wednesday April 19, 2023 @07:08PM (#35634448)

I am really hoping Intel disrupts this status quo. The only thing stopping AMD/Nvidia is anticompetitiveness.

[ Parent | on HN ]

Re:Stability AI Launches the First of Its StableLM Suite of ... by kristianp on Wednesday April 19, 2023 @09:59PM (#35635493)

Does llama.cpp use the npu or metal? I thought it uses NEON instructions? So currently those units aren't even used.

[ Parent | on HN ]

Re:Stability AI Launches the First of Its StableLM Suite of ... by brucethemoose2 on Wednesday April 19, 2023 @10:22PM (#35635612)

I don't think so. M chips just happen to have a really good memory subsystem and good simd performance through accelerate, so the CPU performance is pretty good.

Some stable diffusion implementations can use the NPU or GPU, or (experimentally and unsucessfully) both.

[ Parent | on HN ]