measuredlocally.com
Donate
Apple's MacBook Pro with the M5 Pro or M5 Max chip. The figures here are Apple's, reviewers' and Artificial Analysis's, and each one says whose it is. My own M5 Pro with 64GB arrives in late November.
| Figure | What it is | From | Kind |
|---|---|---|---|
| 24.74 minutes | M5 Max MacBook Pro on the same test, measured by Tom's Hardware23 October 2026, in RTX Spark laptop vs M5 Max MacBook Pro: why the AI reviews are wrong | see the story | a reviewer's test |
| 24.74 minutes | 14-inch MacBook Pro, M5 Max, 128GB, on the same test18 October 2026, in RTX Spark laptop vs M5 Max MacBook Pro: the first independent tests | see the story | a reviewer's test |
| Figure | What it is | From | Kind |
|---|---|---|---|
| 614 GB/s | The M5 Max's memory bandwidth, Apple's figure23 October 2026, in RTX Spark laptop vs M5 Max MacBook Pro: why the AI reviews are wrong | apple.com | the maker's claim |
| Figure | What it is | From | Kind |
|---|---|---|---|
| 5.84 minutes | How far behind the Mac finished23 October 2026, in RTX Spark laptop vs M5 Max MacBook Pro: why the AI reviews are wrong | 24.74 minus 18.9 | my own maths |
| 5.84 minutes | How far ahead the laptop finished18 October 2026, in RTX Spark laptop vs M5 Max MacBook Pro: the first independent tests | 24.74 minus 18.9 | my own maths |
Several reviews put Apple's M5 MacBook Pro ahead of NVIDIA's new RTX Spark laptops for AI, and I think they're wrong. On the one test that replays real AI agent work, the RTX Spark laptop finished in 18.9 minutes and the M5 Max MacBook Pro took 24.74.
The first independent tests of NVIDIA's RTX Spark laptop are out, and on a long AI agent test it finished almost six minutes ahead of Apple's M5 Max MacBook Pro. I put those figures next to my own machines.