• tomatol@lemmy.worldOP
    link
    fedilink
    English
    arrow-up
    22
    arrow-down
    1
    ·
    edit-2
    1 year ago

    Tldw

    • cpu clock is higher and when it boosts it can consume up to 14w, much more than A16
    • however it is still more efficient than the A16
    • gets slightly hotter
    • apple remains ahead of Qualcomm in cpu
    • performance core is around 50% faster than qualcomm 8 gen 2
    • efficiency cores are 2 to 4 times faster than qualcomm 8 gen 2
    • on the gpu side, qualcomm’s 8gen2 from last year beats the iPhone by a small margin using slightly more power
    • GPU performance is around 10% better than the Steam Deck in Resident Evil but it can’t sustain it. (Thermal throttling)
    • Ray tracing performance better than 8gen2
    • conclusion: good SoC but looks like some software tuning might be needed because the cpu boosting to 14w reduces the battery life in the pro models compared to iPhone 14.

    I’m amazed how they managed to get hold of the new iPhones and run all these tests in such a small time!!

    In case you don’t know they also have an English channel here https://youtube.com/@Geekerwan but they tend to post videos later on.

    • thejml@lemm.ee
      link
      fedilink
      English
      arrow-up
      1
      ·
      1 year ago

      Can I strap a Pentium II Slot 1 cooler to the back and keep that GPU perf?

    • tomatol@lemmy.worldOP
      link
      fedilink
      English
      arrow-up
      8
      ·
      1 year ago

      Oh it’s now been re-uploaded on their Chinese channel. I have updated the link. I love that you can edit links on Lemmy!

    • tomatol@lemmy.worldOP
      link
      fedilink
      English
      arrow-up
      4
      arrow-down
      1
      ·
      1 year ago

      It looks like they might have uploaded it by mistake. I managed to watch the whole thing before it was pulled. Will write up a TLDW

  • HaKeNdoR@lemmy.world
    link
    fedilink
    English
    arrow-up
    4
    arrow-down
    3
    ·
    1 year ago

    https://imgur.com/a/h8jkiS2 Amazing they still not able to pass SD8G2 in GPU, even per watt (…oof). Giving they used 3nm processing for these SOCs, I expected at least a little lead. Maybe a software optimization issue?