Pi CM5 - 16GB (AMD Radeon AI Pro R9700) (as tested by Jeff Geerling): edge AI benchmarks
Key results
| Workload | Best result | Model · precision | Power · perf/W | Grade · date | Source |
|---|---|---|---|---|---|
| LLM generation | 19.59 tok/s batch 1 | gpt-oss 20b as in linked issue (Ollama default tags are 4-bit) · llama.cpp (Vulkan) | 318 W (measured) 0.0615 tok/s/W | C 2026-09-22 (page last edited) | Jeff Geerling ai-benchmarks README (Findings tables) |
Best published result per workload, across model sizes. For like-for-like groups, open the per-workload tables below.
Show data — per-workload tables
object detection
No published object detection results for Pi CM5 - 16GB (AMD Radeon AI Pro R9700) (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 Pi CM5 - 16GB (AMD Radeon AI Pro R9700) (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 Pi CM5 - 16GB (AMD Radeon AI Pro R9700) (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 Pi CM5 - 16GB (AMD Radeon AI Pro R9700) (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 Pi CM5 - 16GB (AMD Radeon AI Pro R9700) (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 Pi CM5 - 16GB (AMD Radeon AI Pro R9700) (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 15–40B · b=1 | 19.59 tok/s | gpt-oss 20b as in linked issue (Ollama default tags are 4-bit) · llama.cpp (Vulkan) | C | 318 W measured | Jeff Geerling ai-benchmarks README (Findings tables) 2026-09-22 (page last edited) | 20 of 22 |
How to read these numbers
Machine-readable: /find.json (see llms.txt); full tables are in “Show data”. Confidence grades: A audited MLPerf closed · B documented third-party · C community · D vendor-published · E estimate. Unknown values are shown as unknown, never estimated. Generated 2026-10-16 from public sources.