14 DAY TRIAL // ARM Holdings announced today Project Trillium consisting of a new suite of ARM IP designed from the offset to bring machine learning to edge devices.
The new ARM IP suite will include scalable processors designed to deliver enhanced neural network and machine learning functionality with a focus on the mobile market, according to ARM, which also revealed the fact that they'll enable a new class of machine learning-equipped devices that feature advanced computing capabilities.
"The rapid acceleration of artificial intelligence into edge devices is placing increased requirements for innovation to address compute while maintaining a power efficient footprint. To meet this demand, Arm is announcing its new ML platform, Project Trillium,” said Rene Haas, president, IP Products Group, Arm.
The company said that the high-performance AI and machine learning capabilities, as well as the advanced scalability and flexibility of its new processors developed as part of Project Trillium are required by new devices and will open the door to the development of more advanced smart devices.
Technical specs of the Arm ML and Arm OD processors
The ARM ML (Machine Learning) processor is capable of delivering more than 4.6 trillion operations per second for mobile computing, as well as between 2x and 4x effective throughput through intelligent data management. They're also energy efficient as ARM design them to deliver more than 3 trillion operations per second per watt.
On the other hand, the ARM OD processor has been specifically engineered to efficiently identify objects and people. It's capable of delivering real-time detection with Full HD (1080p) processing at 60fps and up to 80x the performance of a traditional DSP (digital signal processor). They support virtually unlimited objects per frame.
When combined, the two processors that ARM announced today as part of Project Trillium are capable of delivering a power-efficient and high-performance people recognition/detection solution. For end-users, they'll offer real-time, high-resolution, battery-friendly, and detailed face recognition capabilities for smart devices.