# ONNX Runtime plugin-ep-cuda/v0.1.0 — ONNX Runtime CUDA Plugin EP 0.1.0

- Product: ONNX Runtime (https://whatsnew.fyi/product/onnx-runtime)
- Vendor: Microsoft
- Date: 2026-08-17
- Version: plugin-ep-cuda/v0.1.0
- Original notes: https://github.com/microsoft/onnxruntime/releases/tag/plugin-ep-cuda/v0.1.0
- Permalink: https://whatsnew.fyi/product/onnx-runtime/releases/plugin-ep-cuda-v0.1.0

What's New is an index, not a publisher: every entry below links to the vendor's own release notes, which are the authoritative source. Entries are labelled where they are hand-curated sample data, pre-releases, or drawn from a secondary source such as a developer blog.

Reuse: the summaries, labels and curation here are © What's New. Quote freely with attribution and a link back; wholesale republication of the corpus is not permitted — terms: https://whatsnew.fyi/terms. The vendors' own release notes remain their publishers'.

---

- **added** — Introduce CUDA Plugin EP core as the default CUDA provider implementation
- **added** — Add arena allocation, resource accounting, available-resource reporting, and IOBinding synchronization
- **added** — Add provider options for user compute streams, copy behavior, EP-level unified streams, and external allocators with compatibility back to ONNX Runtime 1.24.4
- **added** — Add CUDA Graph capture and replay with user compute stream support, synchronization controls, and kernel sync stream for scratch allocation
- **added** — Add plugin profiling APIs and memory statistics in profiling output
- **added** — Add NHWC hardening, cuDNN SDPA for MHA/GQA, quantized MoE kernels, block-quantized FP4 MatMul, and native NVFP4 QMoE prefill on SM120
- **added** — Add cuDNN-free ArgMax, ArgMin, and ReduceSum operators
- **added** — Add dedicated build and packaging pipelines and onnxruntime-ep-cuda Python wheel
- **added** — Add NuGet packages for CUDA Plugin EP
- **added** — Add Windows ARM64 packages
- **changed** — Make cuDNN and cuFFT optional runtime dependencies
- **changed** — Improve CUDA build coverage across CUDA 12.8/13, SM90, and SM120 with per-architecture object libraries and quick-build support
- **fixed** — Fix LogSoftmax for plugin deployments
- **removed** — Remove nvrtc link requirement

This is the first release of ONNX Runtime CUDA Plugin EP, providing CUDA execution as a separately packaged plugin execution provider.

These notes cover commits affecting CUDA Plugin EP core code, CMake integration, and its primary build and package pipeline.

Please refer to [QUICK_START.md](https://github.com/microsoft/onnxruntime/blob/main/docs/cuda_plugin_ep/QUICK_START.md) for the usage.

##### Highlights

###### Plugin Runtime

- Introduces the CUDA Plugin EP core and makes it the default CUDA provider implementation ([#27816](https://github.com/microsoft/onnxruntime/pull/27816), [#29544](https://github.com/microsoft/onnxruntime/pull/29544)).
- Adds arena allocation, resource accounting, available-resource reporting, and IOBinding synchronization ([#27931](https://github.com/microsoft/onnxruntime/pull/27931), [#28028](https://github.com/microsoft/onnxruntime/pull/28028), [#28103](https://github.com/microsoft/onnxruntime/pull/28103), [#27919](https://github.com/microsoft/onnxruntime/pull/27919)).
- Adds provider options for user compute streams, copy behavior, EP-level unified streams, and external allocators, with compatibility back to ONNX Runtime 1.24.4 through version-gated callbacks ([#28603](https://github.com/microsoft/onnxruntime/pull/28603), [#28824](https://github.com/microsoft/onnxruntime/pull/28824)).

###### CUDA Graphs and Profiling

- Adds CUDA Graph capture and replay, including user compute stream support, synchronization controls, and a kernel sync stream for scratch allocation ([#27958](https://github.com/microsoft/onnxruntime/pull/27958), [#28002](https://github.com/microsoft/onnxruntime/pull/28002), [#28686](https://github.com/microsoft/onnxruntime/pull/28686), [#29221](https://github.com/microsoft/onnxruntime/pull/29221), [#29244](https://github.com/microsoft/onnxruntime/pull/29244)).
- Adds plugin profiling APIs and memory statistics in profiling output ([#28216](https://github.com/microsoft/onnxruntime/pull/28216), [#29058](https://github.com/microsoft/onnxruntime/pull/29058)).

###### Operators and Dependencies

- Expands model coverage with NHWC hardening, cuDNN SDPA for MHA/GQA, quantized MoE kernels, block-quantized FP4 MatMul, and native NVFP4 QMoE prefill on SM120 ([#28612](https://github.com/microsoft/onnxruntime/pull/28612), [#28849](https://github.com/microsoft/onnxruntime/pull/28849), [#28467](https://github.com/microsoft/onnxruntime/pull/28467), [#29818](https://github.com/microsoft/onnxruntime/pull/29818), [#29824](https://github.com/microsoft/onnxruntime/pull/29824)).
- Adds cuDNN-free ArgMax, ArgMin, and ReduceSum and fixes LogSoftmax for plugin deployments ([#29620](https://github.com/microsoft/onnxruntime/pull/29620)).
- Reduces runtime dependencies by making cuDNN and cuFFT optional and removing the `nvrtc` link requirement ([#29252](https://github.com/microsoft/onnxruntime/pull/29252), [#29808](https://github.com/microsoft/onnxruntime/pull/29808), [#29705](https://github.com/microsoft/onnxruntime/pull/29705)).

###### Build and Packaging

- Adds dedicated build and packaging pipelines, the `onnxruntime-ep-cuda` Python wheel, and NuGet packages ([#27959](https://github.com/microsoft/onnxruntime/pull/27959), [#28152](https://github.com/microsoft/onnxruntime/pull/28152), [#28299](https://github.com/microsoft/onnxruntime/pull/28299), [#28378](https://github.com/microsoft/onnxruntime/pull/28378)).
- Adds Windows ARM64 packages, package-size controls, and aligned package metadata and artifact naming ([#28896](https://github.com/microsoft/onnxruntime/pull/28896), [#31635](https://github.com/microsoft/onnxruntime/pull/31635)).
- Improves CUDA build coverage and reliability across CUDA 12.8/13, SM90, and SM120 while reducing CI build time with per-architecture object libraries and quick-build support ([#29711](https://github.com/microsoft/onnxruntime/pull/29711), [#28736](https://github.com/microsoft/onnxruntime/pull/28736), [#29776](https://github.com/microso

_[Truncated at 4000 characters — full notes: https://github.com/microsoft/onnxruntime/releases/tag/plugin-ep-cuda/v0.1.0]_