‹ BackHN Continuity

Thread

M5 Ultra Mac Studio Review

269 points · 262 comments · piotrgrabowski

  1. simonw · · focus · HN ↗
    The numbers I was most interested in are tucked away in a chart towards the bottom - the speed comparison of the Mac Studios v.s. a RTX 5090:

      Qwen3.8 27B tokens/sec generation speed
    
      Prompt size    8K    64K   128K   256K
      RTX 5090 PC    59    51    44     n/a
      M5 Ultra       48    39    32     24
      M3 Ultra       31    23.5  20     15
    
    A whole bunch more comparison numbers in this section: <a href="https:&#x2F;&#x2F;www.macstories.net&#x2F;stories&#x2F;m5-ultra-mac-studio-review-the-dream-mac-for-local-ai-agents&#x2F;#mx-pc" rel="nofollow">https:&#x2F;&#x2F;www.macstories.net&#x2F;stories&#x2F;m5-ultra-mac-studio-revie...
    1. jmyeet · · focus · HN ↗
      The selling point of the M5 Ultra Mac Studio is that you can run much larger models that the 5090 can&#x27;t without swapping. NVidia aggressively segments the market on VRAM for this reason. That&#x27;s why a 5090 has an MSRP of ~$2k (but good luck getting one for less than $4k) while a 6000 Pro, which is basically a 5090 with 96GB of RAM has now soared beyond $15k where 3-6 months ago it was more like $10-11k. A 6000 Pro has the same memory bandwidth but slightly more CUDA units (IIRC ~24k vs ~21k).

      This advantage won&#x27;t be apparent with a 27B model. The 256GB MS can probably run the newer Flash models locally, something you can&#x27;t do on a 5090.

      I don&#x27;t think we&#x27;ll get a successor to the 5090 until late 2028, maybe even 2029. I&#x27;m basing this on the launch date of the 5000 series and that we haven&#x27;t got a midcycle refresh yet. Rumor has it the chips are ready but the 3GB RAM modules are 3-4x the price of the 2GB modules used on the current cards.

      Apple should see a Mac Studio major update in 2028. That might even force NVidia&#x27;s hand. But it&#x27;s really impossible to say what the state of the market will be 2-3 years from now. It may have completely crashed. I suspect not however.

      The interesting thing will be when the bandwidth demands start forcing HBM memory onto these home&#x2F;enthusiast solutions.

      1. weee322 · · focus · HN ↗
        openai make a npu google make npu (tpu no mater) amd buy tellas

        every company make his own npu (without xai)

        probaby in 2028 we will have more concurent firm on market place

Open on Hacker News to reply ↗

Unofficial Hacker News client; not affiliated with Y Combinator.