Horovod
6.0#20 in deep learning softwareFree planFree planAPILinuxmacOSSelf-hosted
VisitHorovod vs ONNX Runtime
ONNX Runtime
7.2#4 in deep learning softwareFree planFree planAndroidiOSWebLinux
VisitWhat both publish
| Fact | Horovod | ONNX Runtime |
|---|---|---|
| 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. | It provides optimizations for inference latency, throughput, memory utilization, and binary size. |
| Purpose | Horovod is a distributed deep learning training framework for TensorFlow, Keras, PyTorch and Apache MXNet. | ONNX Runtime is a production-grade engine for accelerating machine-learning training and inference in existing technology stacks. |
| 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. |
Horovod plans
- HorovodFree
ONNX Runtime plans
- Open sourceFree
Pick Horovod if…
Horovod suits Python teams training TensorFlow, Keras, PyTorch or MXNet models across multiple GPUs. It is also relevant for deployments using Spark or other listed runtime environments.
Pick ONNX Runtime if…
ONNX Runtime suits developers deploying machine-learning models across varied devices and hardware accelerators. It may also suit teams seeking local training or browser and mobile inference.