A Denoise time of 35+ seconds sounds about the same as what I get for my 24-megapixel raw files on my MacBook Pro, which is similar to yours in age and configuration. (Number of megapixels is a lot more useful as a reference for comparison than number of megabytes, because the file size can vary a lot depending on compression, bit depth, etc.)
Note that what you reported aren’t general Lightroom Classic processing times, but only Denoise. This is important because Denoise has a specific requirement: How fast it goes is specifically tied to how many GPU cores your Mac has, and that’s it. But general Lightroom Classic performance also depends on the CPU, and in some cases storage speed and amount of Unified Memory. For example, merging panoramas depends a lot more on the CPU and Unified Memory than on the GPU, which is the opposite of what makes Denoise fast.
Although I’ve been generally happy with Lightroom Classic performance on my M1 Pro up to now, new features are more demanding so I’m finally starting to think about upgrading my 2021 14" MacBook Pro too. Here are my thoughts on your question.
Just having an M4 or M5 should generally run Lightroom Classic faster than our current M1 Pro, because each CPU and GPU core has gotten much faster since the M1. But if your #1 concern is Denoise speed specifically, then you’ll want to make GPU performance a high priority in your next Mac. If you want to stick to laptops (MacBook Air and Pro), then:
- The MacBook Air and base model MacBook Pro are the least appealing Mac laptops for running Denoise quickly, because those have the least number of GPU cores (8 to 10). However, the GPUs in the M4/M5 versions of those Macs might still beat our M1 Pros at Denoise just because the individual cores are faster now.
- If you stick with the Pro tier processor like you and I have now, you get 16 or 20 GPU cores depending on the level you pay for. You can expect 20 GPU cores to run Denoise roughly twice as fast as the 10 cores of the base processor tier of the same generation.
- If you upgrade to the Max processor, you get 32 or 40 GPU cores. This is as far as you can go with a Mac laptop, so if you pay the high price for the 40-GPU-core Max you might expect Denoise to run about 4x faster than the 10-core GPU in the base processor of the same generation.
You could get 8x the base GPU cores if you went up to the Ultra, but the only way to do that is with the most expensive Mac Studio desktop.
The number of CPU cores and the amount of Unified Memory won’t meaningfully affect speed or Denoise processing. Also, the Apple Neural Engine and the new neural cores in the M5 are not known to help Denoise at this time, but that might change.
The following screen shot is from the
ArtIsRight video on YouTube linked here, and illustrates these points: You can see that the fastest Denoise times are turned in by the Macs with the most GPU cores. For the M4 32-core GPU in the MacBook Pro at the top turning in a time of 3 minutes 15 seconds for 10 raw images (a mix of 36, 45, and 60 megapixels), that works out to about 19.5 seconds per image. If they were all 36MP it would take less time, and if they were all 60MP images it would take more time. (My M1 Pro takes over one minute to denoise one 60MP image.)
The chart also shows that although the M5 GPU is about 20% faster than an M4 GPU with the same number of GPU cores in this Denoise test, the new M5 base tier processor is closer to the bottom of the list simply because it only has 10 GPU cores. When Apple finally gets around to releasing the M5 Pro and M5 Max processors (expected in early 2026), their higher number of fast GPU cores should place them closer to the top of this chart. If you have to buy today and Denoise speed is the top priority, the M4 Max or M3 Ultra processor is the way to go. On a budget, you could settle for the 20 GPU cores in the higher spec M4 Pro and that should still be an improvement over our M1 Pros.
(I added the callouts for the number of CPU cores and GPU cores)
View attachment 27621