[zozbot234]

> hitting the SSD constantly to swap layers

Thing is, people in the local llm community are already doing that to run the largest MoE models, using mmap such that spare-RAM-as-cache is managed automatically by the OS. It's a drag on performance to be sure but still somewhat usable, if you're willing to wait for results. And it unlocks these larger models on what's effectively semi-pro if not true consumer hardware. On the enterprise side, high bandwidth NAND Flash is just around the corner and perfectly suited for storing these large read-only model parameters (no wear and tear issues with the NAND storage) while preserving RAM-like throughput.

[spmurrayzzz]

I've tested this myself often (as an aside: I'm in said community, I run 2x RTX Pro 6000 locally, 4x 3090 before that), and I think what you said re: "willing to wait" is probably the difference maker for me.

I can run Minimax 2.1 in 5bpw at 200k context fully offloaded to GPU. The 30-40 tk/s feels like a lifetime for long horizon tasks, especially with subagent delegation etc, but it's still fast enough to be a daily driver.

But that's more or less my cutoff. Whenever I've tested other setups that dip into the single and sub-single digit throughput rates, it becomes maddening and entirely unusable (for me).

[RandomTeaParty]

What is bpw?

[spmurrayzzz]

Bits per weight, its an average precision across all the weights. When you quantize these models, they don't just used a fixed precision size across all model layers/weights. There's a mix and it varies per quant method. This is why you can get bit precision that arent "real" in a strict computing sense.

e.g. A 4-bit quant can have half the attention and feed forward tensors in Q6, and the rest in Q4. Due to how block-scaling works, those k-quant dtypes (specifically for llama.cpp/gguf) have larger bpw than they suggest in their name. Q4 is around ~4.5 bpw, and Q6 is ~6.5.