It's not strange at all. Apple shows us their benchmarks comparing the M2 Ultra against the $7k Xeon from the 2019 Mac Pro. The current Xeon is the 56 core compared in the article, so it's an appropriate comparison.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Mac Pro 2023 = Another Mac Pro 2013 crisis and failure.
- Thread starter sunny5
- Start date
- Sort by reaction score
some preliminary benchmark is just dropped.
Apple's M2 Ultra Seemingly Can't Beat AMD and Intel Rivals
Not enough clocks, not enough cores.www.tomshardware.com
Based on my guesstime from my blender usage perspective, even with fully spec'ed Ultra GPU with Metal renderer...still no bueno against NVDIA/optix renderer, even consumer grade RTX 4000 series cards smoked M2 Ultra. Ballparking the M2 Ultra GPU power, probably sitting around with AMD RX7900XT / HIP which around 3600 median score (blender open data benchmark)
As I just posted on Tom's Hardware:
Shilov is smart and well-informed. So he knows that Primate replaced GB5 with GB6 because they determined GB5 was flawed for MC. And he knows that if we compare their GB6 MC scores, we get the following. So why did Shilov make his comparison based on GB5 only? This doesn't seem like good tech journalism to me, and I'm used to seeing better from Shilov.
GB6 MC
i9-13900KS: 21,661
M2 Ultra: 21,531
i9-13900K: 19,932
AMD Ryzen Threadripper PRO 5995WX: 18,413
[Geekbench's chart doesn't list the Xeon W9-3495X, but it does have the two other comparators in Shilov's article: The i9-13990K, and the 5995WX. The M2 Ultra achieves higher GB6 MC scores than both of them. Indeed, the only one higher, which was not in Shilov's article, is the i9-13900KS. Granted, this is a single M2 Ultra score, not an average. We'll have more accurate info. after the review embargo lifts.]
Sources:
and
Here's a quote from John Poole about why they updated the MC test in GB6. Essentially, they determined the GB5 MC test was too easy, because it sent separate tasks to each core. I.e., the workloads it gave were "embarassingly parallel". This wouldn't properly represent what the processor would need to do when faced with a true MC task, and such tasks are the more typical workload for workstations with high core counts.
"True-to-Life Scaling
The multi-core benchmark tests in Geekbench 6 have also undergone a significant overhaul. Rather than assigning separate tasks to each core, the tests now measure how cores cooperate to complete a shared task. This approach improves the relevance of the multi-core tests and is better suited to measuring heterogeneous core performance. This approach follows the growing trend of incorporating 'performance' and “efficient” cores in desktops and laptops (not just smartphones and tablets)."Geekbench 6 - Geekbench Blog
Last edited:
Literally nothing disappointing about the M2 Mac Pro… it was exactly as expected once the M2 Extreme chip had been nixed… thats not to say its a bad machine or that it won‘t sell… it will sell, there are plenty of customers that would prefer to transition to AS that have 2019 MPs and associated PCIe cards for their workflow that can be used with this machine and cause the least amount of disturbance to their infrastructure, places where the Mac Studio didn’t make sense at all, facilities that are set up around proprietary central high speed storage for example…
I think instead of being slagged off Apple should be commended for getting it over with, the transition is complete, no it wasn’t pretty for the MP but it can now get better and rapidly, no more waiting 7 years for an upgrade… 18 months from now you’ll have 3nm M3 Ultras… maybe a quad chip will be achieved…
The question everyone should ba asking is what happens to the mac pro in 18months or 3 years time and is it worth buying one or waiting… for a lot of people that’ll be a firm no and they will buy Mac Studios, myself included, for others this will work very well and for another group they will shoes to stick with their Intel MPs for another 18 months and see what changes…
The thing is when Apple gave us the trash an in 2013 they forever disrupted the Mac Pro, now that didnt work, but then we got the iMac Pro and that was just a Mac Studio with a screen… and I know loads of people that are still happily using the iMac Pro… even the 2020 iMac is a very powerful machine still… and then we got teh Mac studio and that took care of a huge number of professionals that needed something that wasn’t an iMac… I reckon the current design Mac pro will be the last tower we see from Apple….
I think instead of being slagged off Apple should be commended for getting it over with, the transition is complete, no it wasn’t pretty for the MP but it can now get better and rapidly, no more waiting 7 years for an upgrade… 18 months from now you’ll have 3nm M3 Ultras… maybe a quad chip will be achieved…
The question everyone should ba asking is what happens to the mac pro in 18months or 3 years time and is it worth buying one or waiting… for a lot of people that’ll be a firm no and they will buy Mac Studios, myself included, for others this will work very well and for another group they will shoes to stick with their Intel MPs for another 18 months and see what changes…
The thing is when Apple gave us the trash an in 2013 they forever disrupted the Mac Pro, now that didnt work, but then we got the iMac Pro and that was just a Mac Studio with a screen… and I know loads of people that are still happily using the iMac Pro… even the 2020 iMac is a very powerful machine still… and then we got teh Mac studio and that took care of a huge number of professionals that needed something that wasn’t an iMac… I reckon the current design Mac pro will be the last tower we see from Apple….
Unless Geekbench explains how GB6 MC works, many people will continue to use GB5 because they don't trust the Geekbench 6 results.
Where do they do that? And even if they did, that would be comparing to previous Mac Pro, so that’s users have an idea what to expect.
And no, the 5y-Xeon is not “current” it’s the top SKU from a wide range of Xeons introduced by an Intel recently. I went to HP and configured a Z8 Fury G5 (very well respected high-end workstation brand) workstation with that CPU and 64GB of RAM - HP wants $12k for it (and that’s still with a very crappy GPU). So you are comparing a $7K computer to a $12K computer. For $7K HP gives you a 24-core w7-3455, which is slower than M2 Ultra - and that’s still with a very slow Nvidia T400 GPU.
The Mac Pro is by no means some amazing workstation. But it’s a very solid baseline performer as far as workstations go. PC workstations eclipse it very quickly if you want more oomph and have a bigger budget. But for $7-8k it’s hard to beat, especially if you need a lot of GPU RAM.
Or you can use gaming parts and get the same or better performance for $4k, but that goes for any workstation and does give you large amounts of GPU RAM.
I am not sure I agree with you here. I am very find of GB6’s changes to measuring MC, but it is driven by the desire to have a better estimate for usual consumer applications. Embarrassingly parallel workloads are more common in workstation market, so GB5 can be an adequate MC benchmark.
I’m not disappointed in the new Mac Pro — it is pretty much what I expected. Apple’s direction with Apple Silicon has been to focus primarily on power efficiency and the laptops, which is the core of the Mac’s business. The high-end processing area has limited synergies with the rest of their business, since they are not in server or data center hardware.
A Mac Pro with expandability via PCIe slots is mostly for the niche customers who have these needs, and I think they made the point very well that the base capabilities of the M2 Ultra were very well suited to high-end video processing. It obviously isn’t a focus of the Mac group to expand further into these high-end markets.
We may see more of a push when Apple’s hardware ray tracing makes it into the M-series processors, it depends on how good it ends up being.
A Mac Pro with expandability via PCIe slots is mostly for the niche customers who have these needs, and I think they made the point very well that the base capabilities of the M2 Ultra were very well suited to high-end video processing. It obviously isn’t a focus of the Mac group to expand further into these high-end markets.
We may see more of a push when Apple’s hardware ray tracing makes it into the M-series processors, it depends on how good it ends up being.
I thought the main market for workstations like the M2 Ultra Mac Pro and Studio were creatives doing heavy-duty photo/video/audio processing; and those programs are typically multi-threaded. Are you saying that, even though they're multi-threaded, their tasks are embarassingly parallel (not requiring communication between cores), such that GB5 is a better MC benchmark for them than GB6?
If so, what would be an example of a multi-threaded desktop application that is "cooperatively parallel" (requiring active cooperation/communication among the cores)?
Certainly it's not typical consumer applications (like MS Office and Acrobat) since those are, by contrast, mostly single-threaded. Indeed I would argue that, for most consumers, once you get beyond 4-6 performance cores, MC speed is irrelevant, since they're rarely going to saturate all the cores.
Last edited:
The use cases you describe are indeed great examples for embarrassingly parallel work, so much in fact, that they are often done on the GPU these days. Of course, it will all depend on what you want to do exactly.
Complex scientific simulations come to mind, software development (at least partially, as there is often processing asymmetry)…
I think for a workstation, smaller amounts of faster cores is always better, as it gives you best flexibility. M2 Ultra is decently positioned here.
Apple is doomed™
Yes.... without the Mac Pro as a "halo product" it is really doomed!
The Mac Pro is a "halo product" as much as a mainframe.
Indeed. In fact, Apple itself no longer exists, but the illusion is maintained by a cabal consisting of the Illuminati, the Trilateral Commission, and the Shriners (yes, those guys in the go-karts with the funny hats—who would suspect them?). What few realize is that Steve Jobs is still alive. His "death" was an elaborate scheme so he could take Apple underground and re-establish it as a black ops defense contractor, and a fake Apple imitator was left in its place. That supposedly fancy Apple Silicon chip? Just a repurposed SnapDragon with firmware designed to fool the tech journalists.Apple is doomed™
Jony Ive tried to expose all this by designing product so absurdly non-functional that things would fall apart, but they canned him when they realized what he was doing. And Tim Cook is a Ferengi in a human skin suit. No wonder the fake company is doing so well.
They could have kept this going for at least another few years, but not after releasing that Mac Pro...
Last edited:
No, the base price could have been the same as the 2019 Mac Pro, starting out with the 12 core w5-3425. It would also have PCIe 5.0.
With the extra R&D needed and increased BoM for an ultra low volume Mac that likely has far less than 100,000 units annually...
I highly doubt that the price point would remain the same as the base 2019 Mac Pro. It having PCIe 5.0 would increase the price even further.
And even if say in a different timeline it did, it will not sell more than the 4 years of the 2019 Mac Pro. It would likely still sell far less because use case for almost all Mac workstation users are satisfied by the Mac Studio whose M2 Max base is still $4k cheaper than the base 2019 Mac Pro.
Base Ultra chip of both would be $2k apart.
No one wants to pay for features they'll never use much less subsidize the economies of scale for others if they have no business with the I/O or PSU of the Mac Pro.
Power Macs and Mac Pros in the past sold that well because all other form factors had a lower than "good enough" raw performance and I/O was that bad.
Since as early as 2006 laptops and other form factors became good enough.
You are right, thanks for clearing this up.
Still, I don't think it invalidates my argument. Apple's product page is just an info for potential users showing how much performance they can expect when upgrading from a previous Mac Pro. It's still very different than arguing that Apple workstation is a dud because their $7000 computer is a slower than an Intel $13000 computer.
I am grateful that he shared his thoughts with the Mac Studio in relation to the Mac Pro.
He has the most salient point as a business owner in relation to the Mac Pro being a business expense.
He isn't that hung up about the cost of upgrades as he has contracts that paid for it outright.
He has a schedule on when his new replacement will appear and if IIRC it is after 7 years.
This fits into the probable support timeline of 8 years for his 2019 Mac Pro.
By 2026 it is likely a 2nm M4 Ultra/Extreme with 768GB unified memory would be available to match his business requirements.
The Intel Mac he has now fits his business case and he sees little reason to replace until Apple levels up the unified memory.
He appears to have no interest in migrating to Windows, much less Linux, for upgradeability. He is the customer that Apple is focusing on as they're the most popular use case and least expensive to support due to economies of scale.
Last edited:
I said it few months back, don’t expect Apple to give external memory on Mac Pro. It’s too niche of a product to deviate from unified RAM and build separate designs for Jacky RAM modules. Hopefully with M3/M4, Apple can figure out an M3 or M4 extreme with 512 GB RAM.
It’s probably going to be that way going forward till Apple can get to 512/ 1 TB GB unified RAM.
I wondered about that too so I checked their FAQ about how they determined share and it's all browser page views. I would imagine that it's how well the browsers are handling privacy concerns, or how fast they update their algorithms for new browser versions. They, of course, don't tell why. I wouldn't trust them for any point in time measurements, only long term trends.
don't disagree, it's likely good business sense for a number of reasons
but apple is not going to likely be making the fastest desktop/workstation computers on absolute performance.
Intel chips and amd are more than sufficient as well if you want to go down the road and likely less expensive for similar absolute performance. Apple Silicon can't do it all (cheap fast cool upgradeable) . But since power consumption and portability are important it's a great innovation. Apple will just not have the fastest desktops/workstations in an absolute sense and certainly not for the money.
Workstations with PCIe slots are heading towards a niche like mainframes.
Same with upgradeability. There is little demand for it. Check your typical consumer laptop at ~$799 price points. If ever you find one it typically has soldered RAM and only 1 SODIMM slot vacant. That slot may never see an upgrade. Same with the the SSD. It may be present but the end users will prefer to replace it as a whole 5-6 years later.
There is also a key difference in business models: Apple is a system's vendor. Meaning that they sell the finished product, not just the processors. So they can use several parts from the vertical process to subsidize others. In this case, Apple can afford to make very good SoCs because they don't sell those chips elsewhere, meaning that they are not as pressured to make them "cheap" in terms of area for example. Since they're going to recoup the profit from elsewhere in the product.
In contrast; AMD and Intel sell their processors to OEMs, so they only get profit from the processor not the finished system. So they have to prioritize cost, by optimizing their designs for Area first and then focus on power. This is why both AMD and Intel use smaller cores, which allows them for smaller dies. But which have to be clocked faster in order to compete in performance, unfortunately that also increases power.
This is probably they key difference; Apple can afford the larger design that is more power efficient for the same performance. Whereas AMD/Intel have to aim for the smaller design that is less power efficient for the same performance.
Last edited:
Workstation chips and GPUs are actually more expensive. You can get a cheaper PC with gaming hardware and similar absolute performance, but gaming equipment doesn’t have GPUs with high memory capacity. If you are looking for raw CPU/GPU performance and only work with simple workloads gaming PC is a good candidate.