X-ScaleAI is a high performance solution for distributed training for your complex AI problems.
Interested in a free trial?
- Exploiting HPC Technologies for Deep Learning
on x86 and OpenPOWER Platforms
- Fine-tuned CUDA-Aware MPI library for GPU clusters
- Fine-tuned MPI library for CPU-only systems
- Distributed Training with TensorFlow, Pytorch, and MXNet using Horovod
- “Out of the box” optimal performance on OpenPOWER (POWER9) + GPU platforms, including #4 Summit system
- “Out of the box” optimal performance on CPU-only platforms, including #16 Frontera system
- Tested with ResNet-50 Training using TensorFlow benchmark on #16 Frontera using up to 2,048 nodes and on Summit using up to 1,536 Volta GPUs
- Simple installation and execution in one command
- Coming soon: support for more system configurations
X-ScaleAI offers a one-command installation process.
Installing the X-Scale AI Package
X-ScaleAI also offers a simple run command.
Sample Run using X-Scale AI Run
Deep Introspect (DI) User Interface
Sample Xscale-AI-Run profiling information displayed using the Deep Introspect Interface.