Horovod vs ONNX Runtime
Horovod
horovod.aiONNX Runtime
onnxruntime.aiSide by side
| Figure | Horovod | ONNX Runtime |
|---|---|---|
| Price from | Free | Free |
| Free plan | Yes | Yes |
| Free trial | — | — |
| Open source | — | — |
| Web | — | Yes |
| Windows | — | Yes |
| macOS | Yes | Yes |
| Linux | Yes | Yes |
| iOS | — | Yes |
| Android | — | Yes |
| API | Yes | — |
| Self-hosted | Yes | Yes |
| Plans published | 1 | 1 |
| Spec score · Deep Learning Software | 8.1 (#11) | 8.4 (#6) |
| Deployment | Horovod can be installed on-premise or run on AWS, Azure and Databricks. | Inference is described for cloud servers, edge and mobile devices, and web browsers. |
| Performance | Tensor Fusion batches small allreduce operations and interleaves communication with computation. | The runtime optimizes latency, throughput, memory utilization, and binary size across CPU, GPU, and NPU hardware. |
| Purpose | Horovod is a distributed deep learning training framework for TensorFlow, Keras, PyTorch and Apache MXNet. | ONNX Runtime is a cross-platform machine-learning model accelerator with interfaces for hardware-specific libraries. |
| Support | The project directs users to documentation, troubleshooting guidance and a ticket submission link for unresolved issues. | Documentation questions are directed to issue filing, and the project invites users to report bugs, suggest features, and submit pull requests on GitHub. |
Plans
In their own words
ONNX Runtime
ONNX Runtime is a free engine for machine-learning inference and training within existing technology stacks. It can run models from frameworks including PyTorch, TensorFlow/Keras, TFLite, and scikit-learn. For inference, it optimizes latency, throughput, memory use, and binary size through graph transformations, accelerator-aware partitioning, and optimized computation kernels. Its Execution Providers connect ONNX models to hardware acceleration libraries for CPUs, GPUs, FPGAs, and specialized NPUs. Listed options include NVIDIA CUDA and TensorRT, Intel OpenVINO, DirectML, Qualcomm QNN, Android NNAPI, Apple CoreML, and WebGPU. Deployment spans cloud servers, edge and mobile devices, and browsers; its web runtime supports browser models and its mobile runtime supports Android and iOS. Supported languages include Python, C#, C++, Java, JavaScript, and Rust. Developers can create smaller web or mobile packages by including only the operators and opsets their models need. The open-source runtime is available on Android, iOS, Linux, macOS, self-hosted environments, web, and Windows.