4 ms·

With Apple Silicone, I think Apple is in the perfect place to take advantage of the AI craze to sell hardware with something like "Neural Engine V2". If they ca

With Apple Silicone, I think Apple is in the perfect place to take advantage of the AI craze to sell hardware with something like "Neural Engine V2". If they can figure out how to run AI workloads at fraction of the cost of what is costs on current hardware, I can see there being a huge market for their Mac's and they could even bring back XServe and become the predominant player in the "Server Hardware for AI" space.

smoldesu 3y ago
> If they can figure out how to run AI workloads at fraction of the cost of what is costs on current hardware https://en.wikipedia.org/wiki/Laconic_phrase https://en.wikipedia.org/wiki/Laconic_phrase It's discussed quite a bit on this site, but I wouldn't expect Apple to pull ahead in the GPU department. They really lit a fire under Intel and AMD's ass with the M1's IPC, but Nvidia laughed their way to the bank with every M1 upgrade. Even the 5nm M1 Ultra struggled to keep up with Nvidia's 10nm, bog-cheap 30-series cards. In the datacenter it's even more of a blowout, Apple would have to invest in something competitive with CUDA to turn heads. That's no small feat, and I don't think it's possible with an overnight API launch. It takes time and integration into the industry, something Apple wasn't patient enough for (see: Xserve).
- pazimzadeh 3y ago
  What about in terms of performance per watt, rather than pure performance?
  - smoldesu 3y ago
    It's still not favorable, even compared to last-gen (30XX series) hardware. Here's the OpenCL scores for multiple different GPUs, including the Ultra: https://browser.geekbench.com/opencl-benchmarks https://browser.geekbench.com/opencl-benchmarks The M1 Ultra is a 200w chip, fabricated with 5nm silicon. It's being outclassed by the RTX 3060, a cheap 10nm card that draws ~170w max. In Nvidia's 40-series cards (on 4nm), the M1 Ultra's performance profile is most comparable to a laptop-class RTX 4060 that draws less than 115 watts. There's a story in the total package draw to be made here, and it does weigh slightly in Apple's favor. Overall though, even with last-gen cards it's clear that there's a massive performance-per-watt lead in Nvidia's favor right now. Which is impressive, considering how they've been stuck with second-class silicon when the M1 Ultra debuted.

smoldesu 3y ago

> If they can figure out how to run AI workloads at fraction of the cost of what is costs on current hardware https://en.wikipedia.org/wiki/Laconic_phrase https://en.wikipedia.org/wiki/Laconic_phrase It's discussed quite a bit on this site, but I wouldn't expect Apple to pull ahead in the GPU department. They really lit a fire under Intel and AMD's ass with the M1's IPC, but Nvidia laughed their way to the bank with every M1 upgrade. Even the 5nm M1 Ultra struggled to keep up with Nvidia's 10nm, bog-cheap 30-series cards. In the datacenter it's even more of a blowout, Apple would have to invest in something competitive with CUDA to turn heads. That's no small feat, and I don't think it's possible with an overnight API launch. It takes time and integration into the industry, something Apple wasn't patient enough for (see: Xserve).

pazimzadeh 3y ago
What about in terms of performance per watt, rather than pure performance?
- smoldesu 3y ago
  It's still not favorable, even compared to last-gen (30XX series) hardware. Here's the OpenCL scores for multiple different GPUs, including the Ultra: https://browser.geekbench.com/opencl-benchmarks https://browser.geekbench.com/opencl-benchmarks The M1 Ultra is a 200w chip, fabricated with 5nm silicon. It's being outclassed by the RTX 3060, a cheap 10nm card that draws ~170w max. In Nvidia's 40-series cards (on 4nm), the M1 Ultra's performance profile is most comparable to a laptop-class RTX 4060 that draws less than 115 watts. There's a story in the total package draw to be made here, and it does weigh slightly in Apple's favor. Overall though, even with last-gen cards it's clear that there's a massive performance-per-watt lead in Nvidia's favor right now. Which is impressive, considering how they've been stuck with second-class silicon when the M1 Ultra debuted.