# Torch Xpu Bench

> Benchmark a Hugging Face model on an Intel GPU through pure PyTorch + Transformers, single-process, no HTTP server. Measures generate() throughput in tokens/sec, time-to-first-token, decode-step latency, and peak XPU memory. Also covers diffusion and encoder-only models via references/non-llm-snippets.md. Use after model-can-it-fit to validate predicted memory against torch.xpu.maxmemoryallocated().

## Facts
- Page: https://tashan.sh/capability/plugin-intel-gpu-ai-skills-torch-xpu-bench
- tashan id: plugin:intel/gpu-ai-skills/torch-xpu-bench
- Source: https://github.com/intel/gpu-ai-skills
- Type: plugin
- Category: other
- tashan score: 45.0 / 100
- Adoption: 7.0
- Upkeep: 100.0
- Freshness: 99.0
- Evidence coverage: 84% of the inputs this score can use
- Health: active
- Instruction depth: not yet graded
- License: Apache-2.0
- Official: no

## Install

```sh
/plugin marketplace add intel/gpu-ai-skills
/plugin install torch-xpu-bench@intel-model-skillpack
```

## Security audit
Not scanned. We audit npm-published capabilities; this one has no npm package we can resolve, or has not reached the queue. This is not a clean bill of health.

---
Measured 2026-09-13 by tashan (https://tashan.sh) from public evidence. Scorer s5.