NVIDIA combining AI architecture with Arm machine learning platform
Two leaders in the machine learning/deep learning field have joined forces to push artificial intelligence (AI) into the quickly growing ecosystem of devices that comprise the internet of things (IoT). Announced during the recent GPU Technology Conference in Silicon Valley, NVIDIA and Arm, a subsidiary of conglomerate SoftBank, will work together to enable what Arm EVP Rene Haas described as a “vision of connecting a trillion IoT devices.”
NVIDIA’s open source Deep Learning Accelerator (NVDLA) architecture allows developers access to AI tools to incorporate deep learning inferences into designs. Deep learning inference is the result of what’s called training–for image recognition, AI is trained by viewing lots and lots of images; once trained, AI can be shown a new image and correctly infer what it is. NVDLA will be integrated with Arm’s Project Trillium machine learning platform. According to the companies, the goal is to simplify AI integration for IoT chip companies.
Karl Freund, lead analyst for deep learning at Moor Insights & Strategy, called the arrangement “a win/win for IoT, mobile and embedded chip companies looking to design accelerated AI inferencing solutions. NVIDIA is the clear leader in [machine learning] training and Arm is the leader in IoT end points, so it makes a lot of sense for them to partner on IP.”
Back to this concept of inferencing. Deepu Talla, NVIDIA vice president and general manager of autonomous machines, predicted it “will become a core capability of every IoT device in the future.” To give a hypothetical example, an AI-equipped security camera could recognize known faces and allow entrance to a home or businesses, while flagging unknown faces to trigger a manual admittance process. This idea of moving AI to the edge, particularly in support of the IoT, is an emerging area that’s making waves in the silicon community. Qualcomm Technologies is also a major player in the space; to learn more, check out this interview with Qualcomm’s Gary Brotman.
“Accelerating AI at the edge is critical in enabling Arm’s vision of connecting a trillion IoT devices,” Haas said, noting the partnership will help realize that vision “as our entire ecosystem will immediately benefit from the expertise and capabilities our two companies bring in AI and IoT.”
