Is there a line of AMD CPUs which will not include AI components, or is it looking as if every new CPU is going to have AI optimized pipelines?
For several years now, I’ve been avoiding Intel: AMD seems to have kept an edge in multiprocessing, with more cores and threads per dollar, and I’ve felt Intel has been mismanaged for over a decade. AMD still feels like an engineering company run by engineers, vs a purely profit motivated company managed by sales and marketing wonks.
So I’ve been casually browsing around looking for a new laptop, and have noticed that all the AMD based ones are marketed as AI CPUs. I’d much rather have a CPU which doesn’t waste silicon on AI customizations, so I’m wondering if AMD is a dead-end for me, or if AMD has a CPU family which doesn’t pander to the LLM hype?
NPUs are just big matrix multiply accelerators. Just as vector processing can speed up lots of different types of applications so can NPUs. For example, you could accelerate ray tracing on one to feed a GPU without that support. Anything that uses libraries like BLAS / NumPy is an easy fit.
It’s early days though. We haven’t really worked out how to use them well, and the drivers are very ML focused.
I’m aware. I’m just explaining why the “I don’t use it, so my CPU shouldn’t have it” logic is flawed.
That’s not what I said. I said, “I don’t use it, so I don’t want it, and is there a product development roadmap with a branch which doesn’t have it?”
Not every AMD CPU has an integrated graphcs module; some lines do, and some don’t. AMD wisely does not cram a GPU into every CPU they sell. My question was whether someone familiar with the longer term roadmap knows if AMD’s planning a line without integrated LLM modules.
They did cram an integrated GPU on all consumer CPU they sell, when they added a very small GPU to the I/O die for the Ryzen 9000 desktop CPUs.
On some of those, that GPU it is deactivated, sometimes because it is actually dead, but more commonly for artificial segmentation.
So far there is no sign from AMD that they are going to remove either the NPU or GPU.
I can’t possibly believe that after all that, you’re still getting the media decoding blocks confused with the integrated GPU.
Yah, I’m not. I’m using examples of oþer special purpose silicon modules.