M4 Mac mini (10 core CPU) / 32GB (as tested by Jeff Geerling): edge AI benchmarks
- Independent test Jeff Geerling
Key results
How to read power: results are compared only within the same model. By efficiency ranks only results whose source publishes a power figure; By speed ranks every result. Measured = measured for this workload. Vendor-measured (typical/avg) = the vendor’s own measurement at typical or average load, not an independent test and not a maximum, so fps/W looks better than at max TDP. Chip, Module or Board TDP = the vendor’s published maximum, not measured. Typical power, not max TDP = only a typical figure is published, so fps/W looks better than at full load. Scope says what the watts cover: Accelerator only (host not included), Module, or Whole system. Max of range = top of a published range, so fps/W is a floor (≥). “— not published” = the source gives no power; we never estimate one.
| Workload | Model · precision | Result | Power | Perf/W | Rank on this model | Grade · date | Source |
|---|---|---|---|---|---|---|---|
| LLM generation | llama2:13bOllama default tag (4-bit unless tag says otherwise) · batch 1 | 13.60 tok/s | 29.80 WWhole system Measured | 0.456● Measured · Whole system | not in the default fps/W ranking #2 of 6 by speed | C 2024-11-11 (page last edited) | Jeff Geerling sbc-reviews (AI / LLM inference tables)ai-benchmarks README @ 2f56f19 |
| llama3.1:8bOllama default tag (4-bit unless tag says otherwise) · batch 1 | 20.95 tok/s | 29.40 WWhole system Measured | 0.713● Measured · Whole system | not in the default fps/W ranking #4 of 8 by speed | C 2024-11-11 (page last edited) | Jeff Geerling sbc-reviews (AI / LLM inference tables)ai-benchmarks README @ 2f56f19 | |
| llama3.2:3bOllama default tag (4-bit unless tag says otherwise) · batch 1 | 41.31 tok/s | 30.10 WWhole system Measured | 1.37● Measured · Whole system | not in the default fps/W ranking #13 of 35 by speed | C 2024-11-11 (page last edited) | Jeff Geerling sbc-reviews (AI / LLM inference tables)ai-benchmarks README @ 2f56f19 |
Ranks compare this board only with boards running the same model. Up to three models per workload; open a rank to see the full same-model table.
Show data — per-workload tables
object detection
No published object detection results for M4 Mac mini (10 core CPU) / 32GB (as tested by Jeff Geerling) yet.
We only show cited public figures — nothing is estimated to fill gaps. Requests tell us what to source or test next.
instance segmentation
No published instance segmentation results for M4 Mac mini (10 core CPU) / 32GB (as tested by Jeff Geerling) yet.
We only show cited public figures — nothing is estimated to fill gaps. Requests tell us what to source or test next.
pose estimation
No published pose estimation results for M4 Mac mini (10 core CPU) / 32GB (as tested by Jeff Geerling) yet.
We only show cited public figures — nothing is estimated to fill gaps. Requests tell us what to source or test next.
image classification
No published image classification results for M4 Mac mini (10 core CPU) / 32GB (as tested by Jeff Geerling) yet.
We only show cited public figures — nothing is estimated to fill gaps. Requests tell us what to source or test next.
semantic segmentation
No published semantic segmentation results for M4 Mac mini (10 core CPU) / 32GB (as tested by Jeff Geerling) yet.
We only show cited public figures — nothing is estimated to fill gaps. Requests tell us what to source or test next.
speech-to-text
No published speech-to-text results for M4 Mac mini (10 core CPU) / 32GB (as tested by Jeff Geerling) yet.
We only show cited public figures — nothing is estimated to fill gaps. Requests tell us what to source or test next.
LLM generation
| Like-for-like group | Best result | Model · precision · stack | Grade | Power | Source · date | Rank in group |
|---|---|---|---|---|---|---|
| LLM ≤4B · b=1 | 41.31 tok/s | llama3.2:3b Ollama default tag (4-bit unless tag says otherwise) · Ollama (geerlingguy/ollama-benchmark) | C | 30.10 W ● Measured, Whole system | Jeff Geerling sbc-reviews (AI / LLM inference tables) Older data (2024) | 49 of 81 |
| LLM 5–14B · b=1 | 20.95 tok/s | llama3.1:8b Ollama default tag (4-bit unless tag says otherwise) · Ollama (geerlingguy/ollama-benchmark) | C | 29.40 W ● Measured, Whole system | Jeff Geerling sbc-reviews (AI / LLM inference tables) Older data (2024) | 44 of 66 |
Grades: A audited MLPerf · B third-party · C community · D vendor. How we rank · JSON