Shulou(Shulou.com)11/24 Report--

Thanks to CTOnews.com netizen Coje_He for the clue delivery! CTOnews.com, March 12, according to Google's official blog, the development and deployment of machine learning ML is now affected by decentralized and isolated infrastructure, which may vary depending on framework, hardware, and use cases. This fragmentation limits the speed of developers and hinders the portability, efficiency, and production of the model.

To this end, 12 technology giants Alibaba, Amazon AWS, AMD, Apple, Arm, Cerebras, Google, Graphcore, Hugging Face, Intel, Meta and Nvidia announced a joint OpenXLA project (including XLA, StableHLO and IREE repositories) that enables developers to compile and optimize models from all leading ML frameworks for efficient training and services on a variety of hardware.

According to reports, the OpenXLA project provides a state-of-the-art ML compiler that can be extended in a complex ML infrastructure. This general-purpose compiler can bridge different hardware devices to multiple frameworks in use today (such as TensorFlow, PyTorch), and use OpenXLA to help accelerate the development and delivery of AI to realize the actual potential of AI.

Google says developers using OpenXLA will see significant improvements in training time, throughput, service delays, as well as final time to market and computational costs.

The OpenXLA project has been uploaded to GitHub,CTOnews.com with an access link.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.