Scaling Distributed Machine Learning with In-Network Aggregation
This addresses a key bottleneck in parallel training for ML practitioners, offering a significant performance improvement.
The paper tackles the problem of slow distributed machine learning training by designing SwitchML, a communication primitive that aggregates model updates in programmable network switches, reducing data exchange volume and achieving up to 5.5× speedup on real-world benchmarks.
Training machine learning models in parallel is an increasingly important workload. We accelerate distributed parallel training by designing a communication primitive that uses a programmable switch dataplane to execute a key step of the training process. Our approach, SwitchML, reduces the volume of exchanged data by aggregating the model updates from multiple workers in the network. We co-design the switch processing with the end-host protocols and ML frameworks to provide an efficient solution that speeds up training by up to 5.5$\times$ for a number of real-world benchmark models.