Can a Jetson Orin Nano Super run Phi-3 3.8B?
Yes, usably. Best reported generation speed: 16.9 tokens/s on Jetson Orin Nano Super (Q4, llama.cpp CUDA fbd441c37 (MAXN)), per PicoCluster Claw. Boards tested: Jetson Orin Nano Super.
Phi-3 3.8B speed by board
- Jetson Orin Nano Super: 16.9 tokens/s best (Q4, llama.cpp CUDA fbd441c37 (MAXN)), 1 test
FAQ
Which edge board runs Phi-3 3.8B fastest?
Of the boards tested, Jetson Orin Nano Super is fastest for Phi-3 3.8B: 16.9 tokens/s generation (Q4, llama.cpp CUDA fbd441c37 (MAXN)).
How fast does Phi-3 3.8B run on Jetson Orin Nano Super?
Best reported generation speed for Phi-3 3.8B on Jetson Orin Nano Super is 16.9 tokens/s (Q4, llama.cpp CUDA fbd441c37 (MAXN)), from 1 sourced test. Source: PicoCluster Claw (https://picocluster-claw.com/reference/benchmarks/).
| Board ↕ | Model ↕ | Runtime & setup ↕ | Gen t/s ↕ | Price ₹ (India) ↕ | Source ↕ |
|---|---|---|---|---|---|
| Jetson Orin Nano Super8 GB RAM | Phi-3 3.8B3.8B params · Q4 | llama.cpp CUDA fbd441c37 (MAXN)NVMe 938GB + eMMC | 16.9 ⚠ unusual, check sourceprompt 537.0 t/s | ₹67,980 Buy0.25 t/s per ₹1k | Vendor PicoCluster Claw2026-04-02 |
Each row shows only what its source reported: RAM, quant, prompt speed and setup appear under the main value when published. Rows with full details are listed first, fastest first. Source badges: Community, Vendor, Reproduced; ⚠ marks a speed far from similar results (how we check).