“AI Everywhere means that anybody should be able to apply and use AI,” says Li, explaining the pledge Intel has made to further democratise the use of technologies such as AI. “We provide not only the hardware for AI, but also AI software and solutions for everyone to accelerate their data to insight journey. Software is the bridge between hardware and the millions of developers and billions of users.”
With Intel’s size and scale, the corporation has the unique capability to help build and deliver the full gamut of AI tools and solutions in the market. This can range from language and model optimisations, open-source development tools, to full AI solutions or business partnerships. Whatever a customer needs to get AI deployed in their business, Intel has options to help get them going. “We are working end-to-end,” adds Li. “We want to make it easy for people to build and deploy AI across any industry.”
The innovation journey at Intel
Across more than 20 years with the company, Li has been a part of Intel’s journey as it transformed from a hardware company selling processors for PCs and powering the world’s largest data centres, to the move to the cloud (Intel is among the world’s leading providers of server tech) and its transformation into both a hardware and software provider.
Indeed, for Artificial Intelligence and Intel, there is a hardware and software story. “Our software runs best on Intel hardware, giving a combined advantage to solving customer problems,” says Li. “For hardware, this starts with AI-specific features built into our Xeon CPU processors and extending to our work with Xe GPUs and AI accelerators such as Habana.”
While Intel may be most known as a hardware company, it is going through a transformation to software first, inspired to become the preferred compute platform for the world.
“We have one of the largest teams of software engineers (15,000+) working for any company. There’s a lot of talent inside Intel working to solve people’s problems with software… We are one of the biggest contributors to open-source software, and have been pivotal contributors to several software projects such as Linux Kernel, OpenCV, OpenMP, and OpenStack, among others,” notes Li. “In AI, this extends to the optimisations that we have made and up-streamed to the most popular industry AI frameworks such as TensorFlow, PyTorch, SciKit-Learn, XGBoost, and others. These framework optimisations along with our rich suite of AI tools provide developers with an end-to-end AI software portfolio.”