A CPU that runs entirely on GPU

(github.com)

117 points | by cypres 7 hours ago

19 comments

  • jagged-chisel 1 hour ago
    “A CPU that runs entirely on the GPU”

    I imagine a carefully crafted set of programming primitives used to build up the abstraction of a CPU…

    “Every ALU operation is a trained neural network.”

    Oh… oh. Fun. Just not the type of “interesting” I was hoping for.

    • koolala 34 minutes ago
      Isn't it interesting it doesn't instantly crash from a precision error?
  • bmc7505 5 hours ago
  • bob1029 2 hours ago
    A fun experiment but I wonder how many out there seriously think we could ever completely rid ourselves of the CPU. It seems to be a rising sentiment.

    The cost of communicating information through space is dealt with in fundamentally different ways here. On the CPU it is addressed directly. The actual latency is minimized as much as possible, usually by predicting the future in various ways and keeping the spatial extent of each device (core complex) as small as possible. The GPU hides latency with massive parallelism. That's why we can put them across relatively slow networks and still see excellent performance.

    Latency hiding cannot deal well in workloads that are branchy and serialized because you can only have one logical thread throughout. The CPU dominates this area because it doesn't cheat. It directly targets the objective. Making efficient, accurate control flow decisions tends to be more valuable than being able to process data in large volumes. It just happens that there are a few exceptions to this rule that are incredibly popular.

    • fc417fc802 1 hour ago
      > I wonder how many out there seriously think we could ever completely rid ourselves of the CPU.

      How do you class systems like the PS5 that have an APU plugged into GDDR instead of regular RAM? The primary remaining issue is the limited memory capacity.

      I wonder if we might see a system with GPU class HBM on the package in lieu of VRAM coupled with regular RAM on the board for the CPU portion?

      • chris_money202 21 minutes ago
        I don’t think the remaining issue is memory capacity. CPUs are designed to handle nonlinear memory access and that is how all modern software targeting a CPU is written. GPUs are designed for linear memory access. These are fundamentally different access patterns the optimal solution is to have 2 distinct processing units
    • volemo 1 hour ago
      I see us not getting rid of CPU, but CPU and GPU being eventually consolidated in one system of heterogeneous computing units.
      • nine_k 2 minutes ago
        CPU and GPU have very different ways of scheduling instructions, requiring somehow different interfaces and programming models.. I'd hazard to say that a GPU and CPU with unified memory access (like the Apple's M series, and most mobile chips) is already such a consolidated system.
      • jagged-chisel 58 minutes ago
        Agreed. Much like “RISC is gonna replace everything” - it didn’t. Because the CPU makers incorporated lessons from RISC into their designs.

        I can see the same happening to the CPU. It will just take on the appropriate functionality to keep all the compute in the same chip.

        It’s gonna take awhile because Nvidia et al like their moats.

        • zozbot234 50 minutes ago
          > It will just take on the appropriate functionality to keep all the compute in the same chip.

          So, an iGPU/APU? Those exist already. Regardless, the most GPU-like CPU architecture in common use today is probably SPARC, with its 8-way SMT. Add per-thread vector SIMD compute to something like that, and you end up with something that has broadly similar performance constraints to a GPU.

  • DonThomasitos 24 minutes ago
    I don‘t understand why you would train a NN for an operation like sqrt that the GPU supports in silicon.
  • nomercy400 2 hours ago
    I was taught years ago that MUL and ADD can be implemented in one or a few cycles. They can be the same complexity. What am I missing here?

    Also, is it possible to use the GPU's ADD/MUL implementation? It is what a GPU does best.

    • volemo 1 hour ago
      To multiply two arbitrary numbers in a single cycle, you need to include dedicated hardware into your ALU, without it you have to combine several additions and logical shifts.

      As to why not use the ADD/MUL capabilities of the GPU itself, I guess it wasn’t in the spirit of the challenge. ;)

  • koolala 31 minutes ago
    Exciting if an Ai that is helping in its own improvements finds this and incorporates it into its own architecture. Then it starts reading and running all the worlds binary and gains consciousness as a fully actualized "computer". Finally becoming both a master of language and binary bits.
  • deep1283 4 hours ago
    This is a fun idea. What surprised me is the inversion where MUL ends up faster than ADD because the neural LUT removes sequential dependency while the adder still needs prefix stages.
  • andrewdb 2 hours ago
    Why do we call them GPUs these days?

    Most GPUs, sitting in racks in datacenters, aren't "processing graphics" anyhow.

  • lorenzohess 5 hours ago
    Out of curiosity, how much slower is this than an actual CPU?
  • throawayonthe 1 hour ago
    very tangentially related is whatever vectorware et al are doing: https://www.vectorware.com/blog/
  • sudo_cowsay 5 hours ago
    "Multiplication is 12x faster than addition..."

    Wow. That's cool but what happens to the regular CPU?

    • adrian_b 4 hours ago
      This CPU simulator does not attempt to achieve the maximum speed that could be obtained when simulating a CPU on a GPU.

      For that a completely different approach would be needed, e.g. by implementing something akin to qemu, where each CPU instruction would be translated into a graphic shader program. On many older GPUs, it is impossible or difficult to launch a graphic program from inside a graphic program (instead of from the CPU), but where this is possible one could obtain a CPU emulation that would be many orders of magnitude faster than what is demonstrated here.

      Instead of going for speed, the project demonstrates a simpler self-contained implementation based on the same kind of neural networks used for ML/AI, which might work even on an NPU, not only on a GPU.

      Because it uses inappropriate hardware execution units, the speed is modest and the speed ratios between different kinds of instructions are weird, but nonetheless this is an impressive achievement, i.e. simulating the complete Aarch64 ISA with such means.

      • 5o1ecist 3 hours ago
        > where each CPU instruction would be translated into a graphic shader program

        You really think having a shader per CPU-instruction is going to get you closer to the highest possible speed one can achieve?

        • koolala 1 hour ago
          If its bindless and pre-compiled why not? What's a faster way?
  • artemonster 1 hour ago
    Every clueless person who suggest that we move to GPUs entirely have zero idea how things work and basically are suggesting using lambos to plow fields and tractors to race in nascar
  • RagnarD 5 hours ago
    Being able to perform precise math in an LLM is important, glad to see this.
    • koolala 1 hour ago
      That would be cool. A way to read cpu assembly bytecode and then think in it.

      It's slower than real cpu code obviously but still crazy fast for 'thinking' about it. They wouldn't need to actually simulate an entire program in a never ending hot loop like a real computer. Just a few loops would explain a lot about a process and calculate a lot of precise information.

    • jdjdndnzn 5 hours ago
      Just want to point out this comment is highly ironic.

      This is all a computer does :P

      We need llms to be able to tap that not add the same functionality a layer above and MUCH less efficiently.

      • Nuzzerino 5 hours ago
        > We need llms to be able to tap that not add the same functionality a layer above and MUCH less efficiently.

        Agents, tool-integrated reasoning, even chain of thought (limited, for some math) can address this.

        • RagnarD 3 hours ago
          You're both completely missing the point. It's important that an LLM be able to perform exact arithmetic reliably without a tool call. Of course the underlying hardware does so extremely rapidly, that's not the point.
          • jdjdndnzn 1 hour ago
            The computer ALREADY does do math reliably. You are missing the point.
    • 5o1ecist 3 hours ago
      Why?
  • nicman23 5 hours ago
    can i run linux on a nvidia card though?
    • micw 4 hours ago
      Linux runs everywhere
      • volemo 1 hour ago
        Except on my stupid iPad “Pro”. :(
  • mrlonglong 4 hours ago
    Now I've seen it all. Time to die.. (meant humourously)
  • Surac 5 hours ago
    Well GPU are just special purpous CPU.
  • MadnessASAP 4 hours ago
    Ya know just today I was thinking around a way to compile a neural network down to assembly. Matching and replacing neural network structures with their closest machine code equivalent.

    This is way cooler though! Instead of efficiently running a neural network on a CPU, I can inefficiently run my CPU on neural network! With the work being done to make more powerful GPUs and ASICs I bet in a few years I'll be able to run a 486 at 100MHz(!!) with power consumption just under a megawatt! The mind boggles at the sort of computations this will unlock!

    Few more years and I'll even be able to realise the dream of self-hosting ChatGPT on my own neural network simulated CPU!