Intel AI Platforms Accelerate Microsoft Phi-3 Generative AI Models

Intel worked with Microsoft to enable Phi-3 model support for its central processing units, graphics processing units and Gaudi accelerators. The company also co-designed the accelerator abstraction in DeepSpeed, which is a deep learning optimisation software suite, and extended the automatic tensor parallelism support for Phi-3 and other models on Hugging Face.

The size of Phi-3 models is well-suited to be used for on-device inference and makes lightweight model development like fine-tuning or customisation on AI PCs and edge devices possible. Intel client hardware is supported by software frameworks and tools, including PyTorch and Extension for PyTorch, used for local research and development, and OpenVINO Toolkit, for model deployment and inference.

“Our active collaboration with fellow leaders in the AI software ecosystem, like Microsoft, is key to bringing AI everywhere,” Pallavi Mahajan, general manager of data centre and AI software at Intel, said. “We’re proud to work closely with Microsoft to ensure Intel hardware—spanning data centre, edge and client—actively supports several new Phi-3 models.”