113 pointsby tionis4 hours ago11 comments
  • SwellJoe3 hours ago
    I note the lack of performance information. I can only imagine it's much, much, slower than any other way to run a larger model (including, e.g. using system RAM and streaming some stuff from disk). Consumer networks, even 10gbit ethernet, are slow as hell compared to local RAM and even disks.

    Are we talking 1 token per second for a split model? Less?

    Edit: Found a number. On the models list, Qwen 235B A22B says "MoE 235B/22B, proven at 16 tok/s across 2 nodes". They don't say what the nodes are and what network connection they have, but that's a respectable speed. Not quite comfortable for interactive use, but pretty close.

    • i3862 hours ago
      This was done on my home lab simulating 5ms latency and jitter between machines. Splits work quite well if you your nodes are over WAN at metro latency’s but not super fast on global WAN.

      The idea is that you could take several machines without dedicated RDMA or NVLINK fabric and use them to serve a large model on hardware you own then share it with others.

      I’m currently working on GLM 5.2 on my lab environment with around 10 tok/s on the same split.

      • SwellJoe2 hours ago
        That sounds cool, but it's still pretty meaningless without information about what your home lab looks like. A few DGX Sparks wired up with their fancy super fast network is much different than a few laptops on wifi.
      • zdw2 hours ago
        What hardware (CPU/GPU/memory) and network was used for this? What quantization for GLM 5.2? How much tuning of the split was needed?
        • i3862 hours ago
          The lab features two Mac Studios: an Apple M3 Ultra (32 CPU cores, 80 GPU cores, 256 GB unified memory) and an Apple M1 Ultra (20 CPU cores, 48 GPU cores, 128 GB unified memory), both connected via 1Gbit Ethernet.

          We use a customized Q2 quantization that preserves sensitive tensors at Q8.

          To reduce compute time per layer, we are developing a custom GLM DSA Metal graph.

          While we are not yet approaching MTP, we plan to port our existing MTP implementations from versions 4.7 and 5.1 to 5.2.

          Since GLM's MTP acceptance rate is very high for a single predicted token, we are exploring token prediction techniques to widen the predicted tokens and utilize parallelism for verification.

          • wlesieutrean hour ago
            Equivalent M3 machines no longer for sale from Apple (only up to 96 GB) but can be had on eBay for around $14,000 each
    • woadwarrior013 hours ago
      Perf should be fairly straightforward to ballpark. You'll need to transfer roughly 2 . hidden_size . num_shards bytes over the network per token during autoregressive decoding. And divide that number by chunk size during prefill.
    • 2 hours ago
      undefined
  • dwoosleyan hour ago
    I’ve been curious what a polymorphic botnet that runs one (or multiple) distributed LLMs would be capable of doing. The idea would be to evolve the botnet delivery and payload using the clustered compute of all hosts in the botnet to run LLMs that guides the evolution of various botnet clusters. Bad cluster morphs get caught and cleaned off and bad delivery methods never spread, but the best versions survive to continue to grow.

    What I envisioned for how it works is fairly similar to this, QUIC can actually be more difficult to detect than it seems since it’s very dynamic.

  • i3862 hours ago
    I’m one of the contributors to Mesh LLM and happy to answer any questions. I authored the skippy engine that allows you to split large models across nodes.
    • Creamsicle474 minutes ago
      Hey, this is a super cool project. It's great to see a lot of the IPFS stuff resurfacing again.

      A few questions:

      1.) How does this handle privacy? If you're distributing compute this way then all actors in the compute graph will also know the sequence being computed.

      2.) Any safeguards against malicious actors poisoning model activations?

    • Lerc40 minutes ago
      I have never really delved into kv cache implementation, do they run effectively separate caches per layer?

      If so I can see it all dividing nicely, computation and data size wise and the only slowdown would be in search layer waiting for it's turn. If you pipelined it you could run multiple queries.

      Is anyone doing best-of-n with a n stage pipeline running each query offset by one?

    • DerivativeBS33 minutes ago
      Curious about: does it have fault tolerance if one of the machines goes down mid-inference? Can it dynamically reroute, or does it just retry?
      • i3869 minutes ago
        It can dynamically route. If a machine drops out of split, the topology is recalculated and the request is automatically retried.
    • iotapi3222 hours ago
      This is super impressive, We have a lab with lots of different epycs and different models - to bring them together this way is amazing. Well done!
      • i386an hour ago
        Thank you! AMD is a weak spot in our testing right now. If you’re willing to contribute or let us borrow some compute time, drop in on the Discord.
    • 2 hours ago
      undefined
  • darkpicnic3 hours ago
    Does Mesh LLM encrypt the payload between nodes? Is it possible to read requests from other users?
    • tekacs3 hours ago
      I'm not affiliated, but yes – the main 'point' of iroh is that it's 'dial-a-key', QUIC with encryption based on the keys of the endpoints.
    • metadat2 hours ago
      Just wondering, why do you care about encryption in this context?
      • darkpicnican hour ago
        If payloads to LLMs are being passed around to various nodes, even trusted ones (like friends and family), it gets awkward if you send something very personal. Think sending a medical question to medgemma:27b.
        • oezian hour ago
          Even if transport is encrypted, the LLM computations will always be clear text, right?
  • jmercouris4 hours ago
    I thought about this too, but the throughput over a network is incredibly slow. It’s not usable for interactive use.
    • i3862 hours ago
      That isn’t true. llama RPC is incredibly slow but staged splits in skippy are orders of magnitude faster.
  • turtleyacht4 hours ago
    It sounds like iroh enables distributed compute without having to finangle custom hardware.
  • darkpicnic3 hours ago
    cocompute.ai is already doing this really well.
    • SwellJoe3 hours ago
      Is it? I don't see anything on the website about splitting a model across multiple devices, only about putting local models on the internet, a wholly orthogonal problem (which is already easy with existing tools, since models use an http API).
      • darkpicnic3 hours ago
        Good point. I know cocompute is working on splitting, but it’s not there yet; I was referring to the round-robin delegation within a trusted pool. Mesh LLM looks great too!
    • dnoberon3 hours ago
      Cool, always good to have more in the ecosystem. I love Iroh and hope this continues to succeed.
  • _superposition_2 hours ago
    I just wish I had the hardware to try it out!
  • jing09928an hour ago
    [dead]
  • nttylock4 hours ago
    [flagged]
  • tlongwell-block2 hours ago
    [dead]