Intel has made serious leaps in artificial intelligence and developed its new processor with Baidu. The company, which made several announcements at Baidu’s artificial intelligence conference, offers much higher performance with its new processors.
Intel announced important innovations at the Baidu Create AI conference last Tuesday. Intel announced that it has developed the Nervana Neural Network Processor for Training (NNP-T) in partnership with Baidu, and changed the name of this artificial intelligence processor that it developed last year for the first time. The processor, previously introduced under the name NNP-L 1000, is used in the development of artificial neural networks.
On the software side, Baidu has collaborated with PaddlePaddle’s deep learning system and Intel hopes that these joint efforts with Baidu will have positive results.
- Japan officially approves antibody cocktail
- Oppo unveils new charging solutions technologies
- There will be special upgrades to the PS5 and Xbox Series versions of GTA Online
- 15,000-year-old viruses discovered in Tibetan glaciers
- Apple shows off Watch Series 6’s health and exercise features with a different ad
NNP-T, which is the first processor produced by a major manufacturer to be customized for training neural networks, will be available later this year. Optimized for high-bandwidth RAMs, the processor dissipates the workforce and can withstand very demanding operations. With a 16 nm architecture, the NNP-T performs 3 to 4 times higher than the company’s Lake Crest series chips.
Intel also said Baidu uses Intel’s Optane DC Persistent Memory technology, as well as the company’s various security software. In addition, the company introduced the NNP-I M.2 accelerator for the deep learning structures of Sunny Cove cores.