top of page

Scaletorch Software

Speed Up

Scale your experiments in 4D!

 

  • Unique DAP engine to speed up your deep learning model development 4x – 30x faster on the same GPU capacity. DAP will optimize the key bottlenecks of the training procedure in your PyTorch code using a combination of async execution and compilation. No code changes needed. 

 

  • Distributed training: Divide your training workload across multiple processors while training a huge deep learning model and get 120x speedup. 

​

  • Trials parallelization: Run your script with multiple variations of hyperparameters in parallel and find the best combination faster! Instead of waiting for 10 trials to run on 1 GPU for 10 hours, run 10 GPUs with trials simultaneously for 1 hour. Benefit from productivity. 

​

  • Multi-cloud: If you face issues with cloud providers' quotas all the time, there is a way to increase GPU availability. Register cloud accounts in multiple providers (e.g. AWS, GCP, Azure) and Scaletorch will create a hypercloud based on the all available GPU capacity that you have in all three clouds. No more quota issues!

​

Cost Reduction

By speeding up your calculations, we will naturally reduce your cost. Instead of paying $3 per 100 hours pay $3 per 2 and get same results. 

Optimize even more with spot instances automation - use spot instances like reserved instances, it's 3x cheaper.  

Privacy and Security

  • Data Streaming: Store the data anywhere you want. We don’t require you to save your data on a particular cloud or in Scaletorch. Our platform will handle data streaming automatically. We prefetch data batch by batch, so the waiting time for the training to start is only for the 1st batch. 

​

  • Use your cloud accounts: Stick to something you are used to. Scaletorch will connect to your account and automatically run training jobs faster. All data flow happens inside your existing infrastructure, Scaletorch doesn't save any of your data on our platform. 

Automation

Treat Scaletorch as your new DevOps guy. We will automate all VMs creation and delete it when you no longer need it. Use our Web application, CLI or API to easily integrate to your current workflow. 

Pricing

We charge $0.30 per GPU per hour on top of what your cloud provider charge you. 

bottom of page