Settings

Theme

DeepSeek R1 671B running on 2 M2 Ultras faster than reading speed

twitter.com

96 points by thyrox a year ago · 31 comments

Reader

mythz a year ago

Someone also got the full Q8 R1 running on a $6K PC without a GPU on 2x EPYC with 768GB DDR5 RAM running at 6-8 tok/s [1].

Will be interesting to see the value/performance compared to next gen M4 Ultra's (or Extreme?) vs NVIDIA's new DIGITS [2] when they're released.

[1] https://x.com/carrigmat/status/1884244369907278106

[2] https://www.nvidia.com/en-us/project-digits/

danans a year ago

Check out the power draw metrics. Following the CPU+GPU power consumption, it seems like it averaged 22W for about a minute. Unless I'm missing something, the inference for this example consumed at most .0004 kWh.

That's almost nothing. If these models are capable/functional enough for most day-to-day uses, then useful LLM-based GenAI is already at the "too cheap to meter" stage.

  • danans a year ago

    So it seems like this was actually 7 M2 Ultras, not 2, so .0028 kWh?

teruakohatu a year ago

I am amazed mlx-lm/mlx.distributed works that well on prosumer hardware.

I don't think they specified what they were using for networking, but it was probably Thunderbolt/USB4 networking which can reach 40Gbps.

shihab a year ago

Please note that it’s using pretty aggressive quantization (around 4 bits per weight)

  • doctoboggan a year ago

    Its not that aggressive of a quantization considering that the full model was trained at only 8 bits.

    • shihab a year ago

      That doesn't necessarily mean final weights are 8-bit though. Tensor core ops are usually mixed precision- matmul happens in low precision but accumulation (i.e. final result) is done in much higher precision to reduce error.

      from deepseek v3:

      "For this reason, after careful investigations, we maintain the original precision (e.g., BF16 or FP32) for the following components: the embedding module, the output head, MoE gating modules, normalization operators, and attention operators...To further guarantee numerical stability, we store the master weights, weight gradients, and optimizer states in higher precision. "

    • OisinMoran a year ago

      That's 16x fewer possible values though (and also just 16 possible values full stop). It would be like giving every person on Earth the same shoe size.

rashidae a year ago

This is amazing!! What kind of applications are you considering for this? A part from saving variable costs, fine tuning extensively and security… I’m curious to evaluate this in a financial perspective, as variable costs can be daunting, but not too much “yet”.

I’m hoping NVIDIA comes up with their new consumer computer soon!

iFred a year ago

Complete aside, but I think this is the first time I’ve seen Apple’s internal DNS outside of Apple.

creativenolo 10 months ago

How is this split between two computers?

DrNosferatu a year ago

Heavily quantized…

Still interesting though.

mrcwinn a year ago

Fascinating to read the thinking process of a flush vs a straight in poker. It's circular nonsense that is not at all grounded in reason — it's grounded in the factual memory of the rules of Poker, repeated over and over as it continues to doubt itself and double-check. What nonsense!

How many additional nuclear power plants will need to be built because even these incredibly technical achievements are, under the hood, morons? XD

Keyboard Shortcuts

j
Next item
k
Previous item
o / Enter
Open selected item
?
Show this help
Esc
Close modal / clear selection