Facebook's parent company Meta has launched AI software tools to assist in reducing the switch between Nvidia and AMD chips

Facebook's parent company Meta has launched AI software tools to assist in reducing the switch betwe ...

Metasaid on Monday that it has developed a new set of free software tools for artificial intelligence applications that might allow developers to switch back and forth between different processors.

The new open-source AI platform, built on a PyTorch, is capable of assisting code run up to 12 times faster on Nvidia''s flagship A100 chip or up to four times faster on AMD''s MI250 chip, according to Meta''s findings.

In a blog post, Meta described the speed increase as a sign of the flexibility the software provides.

The software has become a key struggleground for chipmakers seeking to build an ecosystem of developers to utilize their chips. Nvidia''s CUDA platform has been the most popular so far for artificial intelligence work.

The software, according to Meta, is able to easily swap between chips without being locked in once developers customize their code for Nvidia chips.

"In a blog post, the unified GPU back-end support gives deep learning developers more options for hardware vendors, as well as minimal migration costs.

  • Meta Demands Secret Details From Over 100 Firms in FTC Lawsuit: Details

Nvidia and AMD did not respond to requests for clarification immediately.

Meta''s software is used for AI work called inference, which is when machine learning algorithms that previously have been trained on huge amounts of data are called on to make quick judgements, such as deciding whether a photograph is of a cat or a dog.

"This is a multi-platform software experiment," says David Kanter, the founder of MLCommons, an independent group that evaluates AI speed.

Kanter added that this new Meta AI platform would be "good for customer choice."

2022 Thomson Reuters