Intel NUC9V7QNX
Identifier: Intel Core i7-9850H on Intel NUC9V7QNX (Ghost Canyon Compute Element) Form factor: SFF NUC with PCIe slot Status: active
The Ghost Canyon NUC — a 9th-gen Coffee Lake mobile CPU on a “Compute Element” card slotted into a chassis with a PCIe x16 slot for a half-length GPU.
Build & Cost
| Component | Spec | Cost |
|---|---|---|
| NUC9i7QN Extreme Kit | Intel Core i7-9850H (6c/12t), full barebones | $369.98 (2024) |
| RAM | 64 GB DDR4-3200 (Crucial SODIMM) | $119.99 (2024) |
| GPU | MSI GeForce RTX 4060 8 GB — same physical card as the MSI Codex R 4060 | shared (included with Codex R prebuilt) |
| Total invested | $489.97 (4060 cost not double-counted) |
Geekbench AI Runs
| Backend | Component | Single | Half | Quant | Result |
|---|---|---|---|---|---|
| ONNX / DirectML | RTX 4060 8 GB | 13,112 | — | — | (existing template, ID not in source dump) |
| ONNX / CPU | Core i7-9850H | 2,283 | 592 | 2,353 | 29215 |
AI Capability
- Best Single: 13,112 (RTX 4060)
- Best Half / Quant: partial — only the CPU run is fully captured here
- ACI: ~145–250 (estimate; depends on rerunning the GPU benchmark for full numbers) — likely Tier C
This rig is bottlenecked by the 9th-gen CPU and the 8 GB of VRAM. Useful for vision workloads and small LLMs (≤7B) but the Compute Element form factor limits future upgrades. Worth re-benchmarking the GPU to fill in the half/quant numbers.
See Also
- Benchmark Methodology
- Benchmark Index & Rankings
- MSI Codex R — same RTX 4060 GPU on a more modern Intel platform; useful direct comparison for the CPU bottleneck on this rig
- Mac Mini M4 — same SFF form factor, very different architecture (Apple Silicon + Neural Engine); good baseline for “small box, what’s possible in 2024–2026”
- Lenovo ThinkPad X1 Fold — adjacent older-Intel-mobile rig at the bottom of the leaderboard; together they bracket what 9th-gen-era Intel mobile silicon can do for AI workloads