Iel competes in the AI accelerator market, but is also set to iroduce dedicated AI hardware in its PC chips. Given Iel’s domina market share, software vendors have many reasons to support and use Iel’s AI hardware. Powerful AI features could help boost demand for personal computers, creating a multifaceted opportunity for companies like Iel.
According to Hoshio, Iel is looking for opportunities created by artificial ielligence in differe ways. Using the most advanced AI models generally requires powerful AI accelerators installed in cloud data ceers and requires a lot of memory and computation.
The market for such AI accelerators is growing, and Iel has several ways to take advaage of this growth poteial. The company’s specialized Gaudi AI chips are getting a lot of atteion, and in some cases can outperform Nvidia’s market-leading GPUs. Iel also sells data ceer GPUs that can accelerate a wider range of workloads.
But Iel CEO Pat Gelsinger doesn’t believe AI will stay in the data ceer. The company is building AI io all of its products, and the long-term AI opportunity for Iel goes far beyond AI accelerators.
Pervasive artificial ielligence
Beyond dedicated AI accelerators, Iel is also targeting data-ceric AI workloads through its server CPU business. Launched after multiple delays earlier this year, Iel’s Sapphire Rapids chips have built-in AI accelerators aimed at AI inference tasks. While training AI models requires powerful AI chips, Sapphire Rapids provides a way to efficiely run some of these models on data ceer CPUs. Iel estimates that a third of Sapphire Rapids sales are driven by AI use cases.
Iel is competing for data ceer AI workloads, but the company is also betting on bringing AI closer to end users.
Iel is set to launch its Meteor Lake PC CPUs this year, and for that AI is a focal poi. Meteor Lake will feature dedicated AI hardware capable of accelerating a variety of AI workloads and is logical to run directly on a user’s PC. You can’t run a large language model on your laptop, but a lot of lighter AI workloads can be offloaded from the CPU and speed up significaly.
Iel’s AI hardware must be supported by software to work, but software providers have a strong inceive to use Meteor Lake’s AI. For example, a video conferencing software provider can save money by using the clie’s device to power AI features instead of relying on expensive cloud-based AI accelerators.
As Iel brings AI hardware to PCs with Meteor Lake and future chip families, the sheer volume of devices will force developers and software providers to embrace new hardware.
While rival AMD offers dedicated AI hardware in some of its latest laptop chips, Iel’s share of the x86 processor market, excluding gaming consoles and the Iernet of Things, is currely above 80 perce. For software providers, it would make a lot of sense to support Iel’s AI hardware.
Artificial ielligence is a multifaceted opportunity
While Nvidia stands to benefit the most from increased demand for AI accelerators, Iel has more ways to take advaage of long-term AI demand. The company sells its AI accelerators and data ceer processors with AI capabilities, will soon start shipping computer processors with AI hardware, and in its business could produce AI chips for others within the next few years.
There will be many winners in the field of artificial ielligence, and Iel seems to be one of them.
