Recent content by tunafish

  1. T

    NVIDIA discussion [2024]

    What? No. The rules, as defined in antitrust law, basically say that when you have market power (also, as defined in antitrust law), you can no longer do things like deciding who you sell to based on what benefits you, but you will instead have to start treating all customers the same. There's...
  2. T

    NVIDIA discussion [2024]

    Depends, I don't know how large a market share Nintendo has and whether that is considered comparable. If a game console maker wanted to do orders from both AMD and some other source, and AMD punished them for this, they'd quite possibly get sued and lose. Again, a dominant market position is...
  3. T

    NVIDIA discussion [2024]

    The difference is that Nvidia is probably dominant enough in their market that them doing it is illegal. Being a monopolist/having market power is not in itself wrong, it just means that you have to play under different rules. Rewarding brand loyalty (or punishing disloyalty, same thing) is...
  4. T

    Speculation and Rumors: Nvidia Blackwell ...

    Note that you cannot harvest every defective die even if every part of the die is redundant. Faults don't just mean "transistor doesn't work", there are plenty of potential faults that trash the entire chip even if the part it occurred in isn't important. The canonical example is a direct short...
  5. T

    AMD Execution Thread [2023]

    Dispatching work is easy. The problem with treating multiple pieces of silicon as a single GPU is how do you manage memory coherency, when your program touches the same value in two different places, how do you make sure they see the same thing. How would you do coherency? I would expect there...
  6. T

    AMD Execution Thread [2023]

    Yes, just repeating myself here, right now in inference the platform that people are using is PyTorch, with vLLM and running on NVidia hardware. That's where AMD needs to be good at to compete, and other platforms and workloads just matter a lot less. I think that MI300X will do well enough...
  7. T

    AMD Execution Thread [2023]

    Tensorflow is basically obsolete, no-one uses it anymore. Today, you are using a weird in-house thing or PyTorch. Er. The standard is PyTorch, and for inference that usually means using vLLM. Most people run it on NVidia, but that's what's being used, and that's what makes sense to benchmark...
  8. T

    The Intel Execution in [2023]

    The news that they have major outside customers for 18A is much more important than their financial performance of the past quarter. Intel's current state is mostly caused by falling off the leading edge of process tech, if they can catch that train again they can come back in force.
  9. T

    MS leak illustrates new console development cycle

    Ellewood and Brooklin coming late 2024. Based on this slide: It seems like they refresh a bunch of things, but don't add more ram, or a more powerful apu.
  10. T

    MS leak illustrates new console development cycle

    I would like to point out that it's not that simple. ARM has a noticeably weaker memory model than x86. A lot of multithreaded code written for x86 will compile and run unmodified on ARM, for a while... ... Until it crashes with weird heisenbugs caused by the fact that the developer was...
  11. T

    MS leak illustrates new console development cycle

    To be clear, they have made the decision now, as it was marked for late 2022. These slides are just more than a year old, so they had not made that decision back when they were made.
  12. T

    Xbox Series X [XBSX] [Release November 10 2020]

    Do we know that the new models don't support optical media with a separate (optional) player?
  13. T

    AMD: Zen 4, Speculation, Rumours and Discussion

    No. The AMD-provided code in Linux that figures out how many cores a CPU has provides two numbers: How many CCX and how many cores per CCX. Then amount of cores is A*B, and you address cores by CCX number and core number in CCX. Or, to put it more plainly, there is no OS support for any...
  14. T

    Speculation and Rumors: AMD RDNA4 ...

    As I understand it, the big difference is that GCD contains the command processor, while the SED is just arrays of shaders, and command processor is on the base die. This patent is different from the earlier AMD stacked GPU one in that memory controllers are moved form the SEDs to the base die...
  15. T

    AMD RDNA3 Specifications Discussion Thread

    Depends on which spends more transistors, the ALUs or the data movement. If getting operands to the units takes more die area than the ALUs themselves, doubling up on compute might be transistor-efficient to get more use of the data paths, even if the compute sits idle a lot of the time. Just to...
Back
Top