Posted by on
Tags:
Categories: Adobe Amazon facebook Huawei Intel Machine learning Microsoft Nvidia Samsung Sony

Chinese systems colossus #Huawei claims it is developing chips optimized for artificial-intelligence tasks. The silicon will combine an application CPU, a graphics processing unit, and a hardware engine for accelerating #machinelearning algorithms, it’s reported. Technical details are scant, unfortunately. The components will be unveiled later this year, we’re told. Huawei Consumer Business Group CEO Richard Yu hopes the technology will help the electronics giant compete against Google and Apple in the realm of AI processor development. In other words, Huawei wants you using its specialized chips for running artificial intelligence code, be it on a phone or whatever, and not #Apple ‘s nor #Google ‘s. As you probably know, Huawei produces system-on-chips, and has its own fabless semiconductor subsidiary, HiSilicon, located in Shenzhen, China. It has purchased licenses to use various ARM CPU core designs, and developed Kirin chips for use in Huawei’s smartphones and tablet computers. So far, so Apple. The Chinese goliath has an AI research partnership with the University of California at Berkeley, bringing together Berkeley’s AI Research facility (BAIR) and Huawei’s Noah’s Ark Lab. The alliance is looking at deep learning, reinforcement learning, machine learning, natural language processing, and computer vision. #BAIR also works with #Facebook, #Microsoft, #Samsung, #Sony, #Adobe, #Amazon, #Yahoo!Japan, #Nvidia, #Intel, and #Siemens. Noah’s Ark Lab has facilities in Hong Kong and Shenzhen, and its research areas include: Machine learning Data mining Speech and language processing Information and knowledge management Intelligent systems A branch will be opened in Paris, we’re told, and it will do research on intelligent telecommunication networks using machine learning

https://www.theregister.co.uk/2017/07/20/artificially_intelligent_huawei_chippery/

Advertisements

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.