No one will say otherwise: Intel has missed the wave of artificial intelligence that has swept the semiconductor industry over the past decade, despite some efforts in this area. The company is now working intensively to catch up on the issue with Nvidia. With the Gaudi accelerators developed by Habana Labs, with its GPUs for data centers and of course with the eternal Xeon. On the occasion of the WAICF event, which took place in Cannes on February 8th and 9th, Intel launched a large-scale seduction campaign around the contribution of its CPUs to AI.
Position yourself in the inference market
According to Lisa Spelman, corporate VP and GM responsible for Xeon products and solutions, “Intel chips offer better performance for inference.” A not insignificant comment: Inference now represents the majority of the computational needs compared to model training, a sign ensure that these technologies are mature and are now being put into production on a large scale.
It must also be said that the battle in data centers is also against AMD and its Epyc CPUs, whose market shares have increased rapidly over the last five years. If Intel remains dominant with around 75% of the market, it can't be complacent. It therefore makes sense to position yourself as a leader in inference.
A partnership strategy
Lisa Spelman provided an overview of Intel's commercial roadmap, which includes increasing partnerships with customers such as OVHcloud that are launching Xeon-based AI services. “OVH does this to provide local support and respond to certain challenges, particularly with regard to data regulation,” she explains. On this subject, Thierry Souche, technical director of the French provider, indicates that OVHcloud servers will be available later this year with 5th generation Xeon processors.
Another partner, Iterate.ai, is a company focused on scaling generative AI in the enterprise. It has a low-code solution called Interplay. It is currently deployed on around 4,000 edge servers running Xeon and also provides services for vertical industries. Using fast food as an example, Lisa Spelman explores how Iterate.ai and its Interplay solution – running on Xeon – can improve the fast food drive-thru experience.
Iterate.ai uses a large language model combined with speech recognition to help its customers improve the accuracy of their orders. While voice recognition can be used as a secondary source to support on-site employees, it can also help fast food restaurants deal with staffing shortages and peak traffic times.
The Netflix case
Netflix is one of its renowned customers. It uses artificial intelligence in many ways: recommendation engines, suggestions on what to watch next based on your history, etc. The Californian giant also uses AI to improve customer experience and better manage content delivery and network costs. Using Xeon processors would have doubled the performance compared to another solution.
However, the entertainment giant also uses GPUs, with Lisa Spelman admitting, “There is a portion of their AI workloads that they use GPUs for,” before adding, “But our goal is to enable them to make maximum use of their Xeon infrastructure . “Scale it out before you have to invest in an expensive GPU solution.”
Analysis and monitoring of athletes with Xeon and Gaudi
Intel has also partnered with a company called ai.io, which specializes in tracking sports performance, using Xeon CPUs and Gaudi accelerators to improve tracking of Olympic athletes in 3D. The solution is aimed at high-performance athletes and generates complex analyzes that allow them to improve their performance and adapt their training dynamics.
The company is currently working with many athletes preparing for the upcoming Olympics and its goal is to be able to provide them with a 3D response in real time. Because rendering can take a long time, ai.io relies on Gaudi accelerators, “allowing them to provide this capability at a lower total cost of ownership, whereas they previously worked and scaled on GPUs.”
AI PCs, the future of office computing?
Lisa Spelman also recalled the company's investments in developing the AI PC category. This relatively vague marketing term refers to personal computers with better computing capabilities for AI. Intel can count on Movidius' IP, which has been integrated into its CPUs since Meteor Lake. The entire “Wintel” ecosystem is on board, of course Microsoft, but also the major PC manufacturers. Intel estimates that around 70% of PCs will be “AI PCs” by 2028.
Selected to you