Lurkmass's latest activity

  • L
    Lurkmass replied to the thread PS5 Pro *spawn.
    Which is why I get the impression that they're actually overinflating their stated gains with their new upscaler ... Well until you or...
  • L
    Lurkmass replied to the thread PS5 Pro *spawn.
    The device itself (PS5 Pro) is real but the "supposed details" like the disproportionately high RT improvements (upto 4x perf) or the...
  • L
    Lurkmass replied to the thread PS5 Pro *spawn.
    Hmm, I wouldn't be so certain about those things if I were you ... Did the RT claims include upscaling too and is PSSR a software...
  • L
    Personally, I think Work Graphs with the mesh nodes extension may be a closer model to how modern AMD HW works than amplification...
  • L
    Lurkmass replied to the thread Switch 2 Speculation.
    The computational power isn't the only important parameter. The upgraded PS4 SKU will very much come out on top in factors such as fixed...
  • L
    Even if the computations can be efficiently memory localized for specific NPU HW implementations, there's still potentially a lot of...
  • L
    Lurkmass replied to the thread Switch 2 Speculation.
    I think an unfixable hardware problem which easily lead to major circumvention of anti-piracy measures was a strong reason enough alone...
  • L
    Lurkmass replied to the thread Switch 2 Speculation.
    That was due to a hardware design flaw's boot ROM exploit. It was mostly a coincidental decision that the modified Tegra X1 design was...
  • L
    Apple doesn't care about AAA gaming and mobile graphics hardware vendors don't ever see RT itself moving past the tech demo stage and...
  • L
    Lurkmass replied to the thread PS5 Pro *spawn.
    You're forgetting that *hardware performance* is also a function of fixed function unit (rasterizer/depth & blending state...
  • L
    Lurkmass replied to the thread AMD Execution Thread [2024].
    Can they even guarantee testing on an application level to be hardware agnostic when you can implement potentially different code...
  • L
    Lurkmass replied to the thread AMD Execution Thread [2024].
    You can auto-generate certain subsets of CUDA compatible code from HIP code or vis a vis but that doesn't ultimately change the fact...
  • L
    Lurkmass replied to the thread AMD Execution Thread [2024].
    How exactly is it the 'opposite' if ROCm only works on specific AMD HW (CPU/GPU/FPGAs) combinations with Broadcom (collaborative...
  • L
    Lurkmass replied to the thread AMD Execution Thread [2024].
    Indeed I did but I should return that very inquiry to you because how exactly does mentioning OpenCL or 'crowdsourcing' represent a...
  • L
    Lurkmass replied to the thread AMD Execution Thread [2024].
    My question is what exactly do any of these things that he mentioned to do with AMD currently ?
Back
Top