ARM produces two new AI chips for improved facial recognition and object detection

ARM has announced two powerful new processor designs for use in machine-learning hardware, The Verge reports. The ARM Machine Learning (ML) Processor is designed to speed up AI applications including facial recognition, and the ARM Object Detection (OD) Processor is a second-generation design optimized for processing visual data.

“These are new, ground-up designs, not based on existing CPU or GPU architectures,” Jem Davies, ARM’s vice president of machine learning, told The Verge.

The ML chip is expected to mostly be used in mobile devices, and The Verge reports that Davies said ARM is currently in talks with a number of smartphone manufacturers. The OD will likely be used in a wider range of devices, possible including security cameras and drones.

Specialized AI chips are currently only available in high-end smartphones, but ARM believes that will change.
“Our belief from talking to the market is that this will trickle down very, very fast indeed,” Davies said. “In China they’re already talking about putting this in entry-level smartphones from next year.”

Davies also notes the importance of on-device computing, as Google has said it would need to double its server capacity if users reached three minutes a day each of voice search.

The ARM ML is slated for mid-year availability, and the OD is expected to be available by the end of February.

As previously reported, Samsung recently launched a new application processor with neural-network based deep learning for sophisticated image processing, including facial recognition.

Article Topics

AI chips  |  ARM  |  artificial intelligence  |  facial recognition  |  machine learning  |  mobile device