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.
Anyone seen any good M1 Pro vs M5 Pro benchmarking?
- Thread starter AltecX
- Start date
- Sort by reaction score
Ben J.
macrumors 65816
Stageshoot
macrumors regular
The M5 Pro 18cpu/20gpu benchmarks at
CPU side it’s a touch quicker than the m3 Ultra 32 core
GPU wise it’s neck a neck with the M2 Max
CPU side it’s a touch quicker than the m3 Ultra 32 core
GPU wise it’s neck a neck with the M2 Max
Last edited:
Quasselstripper
macrumors 6502
That ms really impressive! Can’t wait to get mine next week
Basic75
macrumors 68030
This compares all generations of Mx Pro chips and some base, Max and Ultra.
Alameda
macrumors 68000
My initial tests show that M5 Pro is between 4 and 5 times faster in DxO PhotoLab exports than an M1 Pro.
In other words, an export that takes about 45 seconds with my M1 Pro takes 9 or 10 seconds with the M5 Pro. I will continue testing.
I will try this more using an external Thunderbolt drive, to control for the effect of the faster storage.
Quasselstripper
macrumors 6502
Ben J.
macrumors 65816
MacBook Pro M1 Pro (16-inch, 2021) Benchmarks
(M1 Pro @ 3.2 GHz (10 CPU cores, 10 GPU cores)
MacBook Pro M5 (14-inch, 2025) Benchmarks
(M5 @ 4.6 GHz (10 CPU cores, 10 GPU cores)
There are two numbers; single-core and multi-core. The multicore performance will be better with more cores. It allows apps to spread the calculations across all the cores, and add the results together at the end. This will be an obvious benefit in tasks like graphics rendering, video and image processing etc., basic number crunching.
In my business, the studio world with audio processing, we were always told to view the single-core performance as most important, because of the strict timing demands of digital audio processing. All audio must be kept in time - you can't just spread the work on all the cores and sum up the results. Well, not most of the time, anyway.
The numbers (CPU only):
M1 Pro
Single-core: 2374
Multi-core: 12259
M5
Single-core: 4227
Multi-core: 17457
It's quite a difference, but I wouldn't call it 'insane'. I think of it as mostly an effect of the big difference in clock frequency, and of course, improved chip design.
But take a look at the intel macs that we used ten years ago, and thought were fabulous machines - and compare them to the M-series. Now that's what I call 'insane':
Mac Benchmarks - Geekbench
browser.geekbench.com
(Edited; the Geekbench numbers I originally posted was not M5 Pro cpu, but M5. Thanks @galad)
Last edited:
Alameda
macrumors 68000
The single core performance is nearly double, and with the faster storage, it IS double. That’s a lot faster!MacBook Pro M1 Pro (16-inch, 2021) Benchmarks
(M1 Pro @ 3.2 GHz (10 CPU cores, 10 GPU cores)
MacBook Pro M5 Pro (14-inch, 2025) Benchmarks
(M5 @ 4.6 GHz (10 CPU cores, 10 GPU cores)
There are two numbers; single-core and multi-core. The multicore performance will be better with more cores. It allows apps to spread the calculations across all the cores, and add the results together at the end. This will be an obvious benefit in tasks like graphics rendering, video and image processing etc., basic number crunching.
In my business, the studio world with audio processing, we were always told to view the single-core performance as most important, because of the strict timing demands of digital audio processing. All audio must be kept in time - you can't just spread the work on all the cores and sum up the results. Well, not most of the time, anyway.
The numbers (CPU only):
M1 Pro
Single-core: 2374
Multi-core: 12259
M5 Pro
Single-core: 4227
Multi-core: 17457
It's quite a difference, but I wouldn't call it 'insane'. I think of it as mostly an effect of the big difference in clock frequency, and of course, improved chip design.
But take a look at the intel macs that we used ten years ago, and thought were fabulous machines - and compare them to the M-series. Now that's what I call 'insane':
Mac Benchmarks - Geekbench
Alameda
macrumors 68000
I ran Geekbench on my M1 Pro and M5 Pro. Here are the results:
M1 Pro:
Single Core: 2,294
Multi-Core: 10,625
GPU (OpenCL): 40,051
M5 Pro (15 Core):
Single Core: 4,301
Multi-Core: 26,203
GPU (OpenCL): 76,195
GPU (Metal): 122,446
Interestingly, I can run the CPU at 4597 or 4695 MHz, but it doesn’t change the Geekbench result.
M1 Pro:
Single Core: 2,294
Multi-Core: 10,625
GPU (OpenCL): 40,051
M5 Pro (15 Core):
Single Core: 4,301
Multi-Core: 26,203
GPU (OpenCL): 76,195
GPU (Metal): 122,446
Interestingly, I can run the CPU at 4597 or 4695 MHz, but it doesn’t change the Geekbench result.
Last edited:
Ben J.
macrumors 65816
Edit; deleted because of factual errors.
Last edited:
Ben J.
macrumors 65816
How does disk speed affect pure cpu performance?
No they weren't. It says M5 in the page you linked. There is a MacBook Pro that ships with the M5 chip, and that's what you listed.
Basic75
macrumors 68030
Ben J.
macrumors 65816
Sorry, my fault. Corrected now. It seems Geekbench don't have numbers for macs w/M5 Pro just yet.
Alameda
macrumors 68000
Yes, you are correct that it does not affect cpu performance but it affects overall computer system performance. This is one reason why my DxO exports were over 4 times faster.
For me, at least, it’s the overall system improvement I care about. One reason I waited for the M5 Pro, for instance, is Thunderbolt 5. I expect that my next SSD drive will support TB 5.
Last edited:
Ben J.
macrumors 65816
Totally different topic. Check if there is already a thread about this. Create a new thread if there isn't. And be prepared to pay up.
Skip that; build your own! (It's cheaper and better and faster...)
(I do not approve of or endorse this man but on this I think he's absolutely right about slapping this together yourself.)