My Photo-Processing Hardware

In October 2021, I purchased the first MacBook Pro with Apple silicon. At the time, I did all my photo processing on an iMac with a 27" Retina display. Because we were already splitting our time between France and the United States, I needed two of these machines, one in each location. The one in the US had a three-terabyte drive, whereas the one in France had a one terabyte internal drive plus external storage; this meant that whenever I moved from France to the United States or back, I had to transfer large amounts of data on a portable drive. In addition, I had a 2017 Intel-based MacBook Pro I used to process photos while traveling.

As my new computer had to replace three machines, I configured it generously. I ordered the machine with an M1 Max with 32 gigabytes of RAM, a ten core CPU, a 32 core GPU, and a four-terabyte SSD. In terms of performance, that was top-of-the-line; I could have gone with 64 gigabytes of RAM and 8 terabytes of SSD, though neither would have increased performance given the software I was using. I sold the other three computers and bought an LG 27" 5K Retina monitor, and, when it became available, an Apple Studio Display. The LG is in Cannes and the Studio Display in the United States, and the only thing that travels back and forth is the computer.

This machine changed everything for me. Not only did I have all my data available at all times without having to fuss with portable drives, but the performance, compared to everything I had experienced before, was incredible, as was battery life: working away from an outlet became more practical, also because running on battery as opposed to AC power did not result in a performance penalty. For four years, this laptop was my only computer, and it basically went wherever I did. I watched Apple Silicon go from M1 to M2, M3, and M4, always with a base, a Pro, and a Max version.

In the fall of 2025, Apple came out with the base version of the MacBook Pro M5. Unusually, the Pro and Max versions of the M5 were not released at the same time. Instead, these models are supposed to come out in early 2026. Nevertheless, I ordered the base M5 with 32 gigabytes of RAM and a 4-terabyte SSD, the machine's maximum configuration. The main reason was that Apple was offering me close to a thousand dollars for my top-of-the-line 2021 M1 Max MacBook Pro.

While I still had the older computer, I ran several tests of things that are typical for my workflow, and I ran the same tests on the new machine. In the following tables, a negative green number in the rightmost column shows how much less time, in percent, the M5 Mac used to complete a given task; contrariwise, a positive red number shows how much more time, in percent, the M5 computer used to complete a task.

1) DxO PureRAW 5

My photo processing workflow starts with PureRAW, software that demosaics my raw files while at the same time applying noise reduction, optimizing lens sharpness, compensating for chromatic aberration and vignetting, and removing lens distortions. This process returns TIFF or DNG files that can then be treated with Photoshop, Lightroom, or any other photo-processing software. The one complaint people usually have is that PureRAW is slow, understandable in view of all the things it does. Let's see how the two Apple silicon machines fared:

Needless to say, I was surprised by these results. It's not as though I ever felt my older machine was slow, but on the new computer, everything seemed to be more fluid and responsive. I ran a few other tests. For example, I manage a website for a non-profit corporation in North Carolina. As part of this, I occasionally receive video footage of lectures that I slightly edit in iMovie, export to file, and then use VideoProc Converter AI to reduce resolution and end up with a smaller file that will stream more easily. Here is how the two computers handled this operation:

2) iMovie and VideoProc Converter AI

The exporting to file from iMovie is the only instance where the new computer was slower than the old one. This struck me as odd, and I repeated the test several times on the M5 machine; unfortunately, I no longer had the M1 Max computer when I noticed this, so I could not repeat this test on it. It is therefore entirely possible that it is a timing or data entry error.

3) Miscellaneous Benchmarks

Please note that, in the table below, we are comparing both scores and times. For comparisons of scores, higher numbers are better, while when comparing times, lower numbers are better. In each case, the M5 Difference column lists the percentage in green when the M5 did better and in red when it did not.

I'm amazed that the 2025 base MacBook Pro handily beats the 2021 top-of-the-line MacBook Pro, and sometimes by a considerable margin. The M5 was particularly fast in those tests that involved AI, such as when running DxO's PureRAW application. Apple said that the graphics cores now have "neural accelerators" built in; I don't have the foggiest notion what this means, but it's true that AI tasks seem to go much faster with the M5. I can't wait to see what kind of performance the M5 Pro and M5 Max will offer. Considering what the Pro and Mac versions of previous generations of Apple silicon did, the results should be nothing short of astounding.

A PDF with a slightly more detailed performance comparison is available here.

Home | Site Info | Family | The Area | Trips | France | Work | Rants | Photography | Odds & Ends

This page was last modified on January 1, 2026
Send feedback about this page to feedback@kiechle.com
https://www.kiechle.com/photography/computer/index.htm
(optimized for Retina display)

All contents © 1999-2026 The Kiechles