And overnight, Intel's Ice Lake is again way behind.
If you're going to be smug why not use a recent Intel chip?
Apple M1 Ultra
201–210 of 855 posts
Re: Apple M1 Ultra
#202Earlier quoted context omitted.
A few points to make... - the shared CPU+GPU RAM doesn't necessarily mean the GPU has to eat up system RAM when in use, because it can share addressing. So whereas the current Mac pro would require two copies of data (CPU+GPU) the new Mac studio can have one. Theoretically. - they do have very significant video decoder blocks. That means that you may use less RAM than without since you can keep frames compressed in f…
Also, the memory model is quite different - with the ultra-fast SSD and ultra-fast on-die RAM. You can get away with significantly less RAM for the same tasks, not just because of de-duplication but because data comes in so quickly from the SSD that paging isn't nearly the hit it is on say an Intel based Mac. I'd expect it to work more like a game console, streaming in content from the SSD to working memory on the fl…
> ultra-fast on-die RAM
The RAM is not on die. It’s just soldered on top of the SoC package.
> All that is to say, where you needed 1.5TB of RAM on a Xeon, the architectural changes on Apple Silicon likely mean you can get away with far less and still wind up performing better.
No, it does not. You might save a bit, but most of what you save is the transfers, because moving data from the CPU to the GPU is just sending a pointer over through the graphics API, instead of needing to actually copy the data over to the GPU’s memory. In the latter case, unless you still need it afterwards you can then drop the buffer from the CPU.
You do have some gains as you move buffer ownership back and forth instead of needing a copy in each physical memory, but if you needed 1.5TB physical before… you won’t really need much less after. You’ll probably save a fraction, possibly even a large one, but not “2/3rd” large, that’s just not sensible.
Re: Apple M1 Ultra
#203RIP Intel
Re: Apple M1 Ultra
#204At what point Apple will put those chips in their servers or sell server chips? It only makes sense for them to take this architecture to the cloud deployments
As more developers move to ARM architecture by buying Macbooks (I did it last year the first time in my life), ARM cloud will grow very fast, and Apple needs growth, so they can't afford not to do it in a few years (probably with M2 architecture they are already thinking of it). Regarding the exact timeline: I don't know :)
They’d have to go all-in on supporting third party OSs like Linux first. Sure, there are projects to bring linux to the M1, but enterprises that buy commercial server hardware will demand 1st party support
Re: Apple M1 Ultra
#205Earlier quoted context omitted.
What on earth are you talking about. That link shows it’s not even half as fast as the 3060, let alone the 3080. In borderlands it got 24 FPS while the 3080 got 52 FPS. How is that on par?
Gaming benchmarks are completely irrelevant when discussing the actual raw power of the GPU. As the other commenter said- look at the actual GPU benchmark in the first graph. Legacy games written for x86 CPUs obviously are going to perform poorly. I recommend you actually read the review and don't just scroll to the worst gaming benchmark you can find.
> Gaming benchmarks are completely irrelevant when discussing the actual raw power of the GPU.
Maybe, but the "raw power" is useless if it can't be exploited.
> Legacy games written for x86 CPUs obviously are going to perform poorly.
Not if they're GPU-bound. Even native performance isn't that impressive
Re: Apple M1 Ultra
#206It sounds like the chip is fast but I wonder, if like other M1 products, the computers built with it will be fairly restricted, like a console, in terms of the hardware they're able to use (not being able to boot off external HDDs, problems with thunderbolt 3 compatibility in peripherals, having to use abstraction layers to run most of the software world indirectly or have specific porting projects dedicated to M1, e…
Maybe this is the same marketing speak as we've seen with the 1600 nits peak, and 1000 nits sustained brightness claim for the new mini led displays. Which later became 500 nits for SDR content, when ambient temperature allows. [0]
I want to see proper benchmarks before getting too exited.
[0] https://www.notebookcheck.net/The-new-MacBook-Pro-14-only-ma...
Re: Apple M1 Ultra
#207> This enables M1 Ultra to behave and be recognized by software as one chip, so developers don’t need to rewrite code to take advantage of its performance. There’s never been anything like it. Since when did the average developer care about how many sockets a mobo has...? Surely you still have to carefully pin processes and reason about memory access patterns if you want maximum performance.
Not sure what is required of a dev, but as an example, Adobe Premiere pro doesn't take any advantage of >1 CPU, at least on Windows. https://www.pugetsystems.com/labs/articles/Should-you-use-a-...
CPU as in core or socket? These days most CPUs are "many-CPU-cores-in-1-socket" and having X CPU cores over 1 or 2 sockets make a small difference, but software does not care about sockets.
Re: Apple M1 Ultra
#208Re: Apple M1 Ultra
#209Is it actually more powerful than a top of the line threadripper[0] or is that not a "personal computer" CPU by this definition? I feel like 64 cores would beat 20 on some workloads even if the 20 were way faster in single core performance. [0] https://www.amd.com/en/products/cpu/amd-ryzen-threadripper-3...
Next-gen Threadripper Pro was also announced today: https://www.tomshardware.com/news/amd-details-ryzen-threadri...
Re: Apple M1 Ultra
#210Earlier quoted context omitted.
>I wonder if they'll just do 4x M1 Max for that. They'll be running out of names for that thing. M1 Ultra II would be lame, so M1 Extreme? M1 Steve?
Steve would be awesome, but a deal with Tesla to use "Plaid" would be perfection.
I would think they could just go to Mel Brooks instead of dealing with Tesla.