So glad someone did this. Have been running big gpus on egpus connected to spare laptops and thinking why not pis.

Datapoints like this really make me reconsider my daily driver. I should be running one of those $300 mini PCs at <20W. With ~flat CPU performance gains, would be fine for the next 10 years. Just remote into my beefy workstation when I actually need to do real work. Browsing the web, watching videos, even playing some games is easily within their wheelhouse.

I've been kicking this around in my head for a while. If I want to run LLMs locally, a decent GPU is really the only important thing. At that point, the question becomes, roughly, what is the cheapest computer to tack on the side of the GPU? Of course, that assumes that everything does in fact work; unlike OP I am barely in a position to understand eg. BAR problems, let alone try to fix them, so what I actually did was build a cheap-ish x86 box with a half-decent GPU and called it a day:) But it still is stuck in my brain: there must be a more efficient way to do this, especially if all you need is just enough computer to shuffle data to and from the GPU and serve that over a network connection.

I really would have liked to see gaming performance, although I realize it might be difficult to find a AAA game that supports ARM. (Forcing the Pi to emulate x86 with FEX doesn't seem entirely fair.)

Really why have the PCI/CPU artifice at all? Apple and Nvidia have the right idea: put the MPP on the same die/package as the CPU.

Now compare batched training performance. Or batched inference.

Of course prefill is going to be GPU bound. You only send a few thousand bytes to it, and don't really ask to return much. But after prefill is done, unless you use batched mode, you aren't really using your GPU for anything more that it's VRAM bandwidth.

Not sure what was unexpected about the multi GPU part.

It's very well known that most LLM frameworks including llama.cpp splits models by layers, which has sequential dependency, and so multi GPU setups are completely stalled unless there are n_gpu users/tasks running in parallel. It's also known that some GPUs are faster in "prompt processing" and some in "token generation" that combining Radeon and NVIDIA does something sometimes. Reportedly the inter-layer transfer sizes are in kilobyte ranges and PCIe x1 is plenty or something.

It takes appropriate backends with "tensor parallel" mode support, which splits the neural network parallel to the direction of flow of data, which also obviously benefit substantially from good node interconnect between GPUs like PCIe x16 or NVlink/Infinity Fabric bridge cables, and/or inter-GPU DMA over PCIe(called GPU P2P or GPUdirect or some lingo like that).

Absent those, I've read somewhere that people can sometimes see GPU utilization spikes walking over GPUs on nvtop-style tools.

Looking for a way to break up tasks for LLMs so that there will be multiple tasks to run concurrently would be interesting, maybe like creating one "manager" and few "delegated engineers" personalities. Or simulating multiple different domains of brain such as speech center, visual cortex, language center, etc. communicating in tokens might be interesting in working around this problem.

Wide interconnects to load and sync do matter, though.