Cerebras CSoft v1.2 Expands Support For PyTorch And TensorFlow
Cerebras' CSoft software update to version 1.2 gives developers using PyTorch and TensorFlow access to their powerful CS-2 systems through Cerebras Cloud.
Created on April 15|Last edited on April 15
Comment
Cerebras provides a number of services revolving around their massive CS-2 system fitted with their Wafer Scale Engine 2 chip, a processor containing 2.6 trillion transistors and 850,000 cores.
This chip is laser-focused on AI computation and training performance and speed, the largest chip for AI processing to date. The CS-2 system is leagues more powerful than any commercially available processor out there.
For extreme deep learning tasks, the machine is controlled through Cerebras' software platform, CSoft. Version 1.2 of CSoft was just released, adding support for PyTorch and TensorFlow, allowing users to easily apply their current projects using these open-sources platforms to the power of the CS-2 through CSoft and even Cerebras' cloud computing services.
Moving your PyTorch and TensorFlow training to the CS-2
Cerebras has made the transition as simple as it could be for users looking to take the power of the CS-2 into their own hands. Just a few simple code additions and compilation through the CSoft software lets you upload your model to the CS-2 through the Cerebras Cloud service effortlessly, giving you blazing fast training speeds unmatched by typical processors setups.
Find out more
Add a comment
Tags: ML News
Iterate on AI agents and models faster. Try Weights & Biases today.