Intel Seeks To Turbocharge The AI PC Movement By Empowering ISVs In New Program

‘We believe this is going to unlock 100 million AI PCs over the next couple of years by 2025, and that’s a scale that we think our competition can’t reach,’ Andrew Marsee, general manager of Intel Partner Alliance, tells CRN.

Intel is unleashing a bevy of engineering, design and marketing resources for more than 100 independent software vendors to deliver on CEO Pat Gelsinger’s promise that AI “will fundamentally transform, reshape and restructure the PC experience.”

The semiconductor giant is providing the resources as part of what it’s calling the “industry’s first” AI PC Acceleration Program. Intel unveiled the program Thursday, less than two months before it releases the first Core Ultra laptop processors, which it has said will “usher in the age of the AI PC.”

[Related: Analysis: Nvidia, AMD Give Partners New AI Selling Points For GPUs In PCs]

The more than 100 ISVs represent Intel’s first wave of engagement for the program, and they include big names like Adobe, Zoom Video Communications and Cisco Systems as well as other ISVs working on AI-accelerated features for PCs, such as Audacity, CyberLink and Topaz Labs.

These ISVs are working on more than 300 AI-accelerated features for PCs within the program, which is part of the larger Intel Partner Alliance program, according to Intel.

Among several statements of support from ISVs, Zoom said it plans to take advantage of the neural processing unit (NPU) inside Intel’s upcoming Core Ultra processors, code-named Meteor Lake, to handle image and audio AI workloads to improve the videoconferencing experience.

“Intel’s adoption of the NPU across their road map means that the Zoom experience will benefit from improved power efficiency for AI workloads starting with Meteor Lake platforms soon with even more to come in the future,” said Eric Yu, head of hardware partnerships at Zoom, in a statement.

The program was unveiled a month after Intel detailed its vision of the AI PC at the third annual Intel Innovation event, where the company showcased AI-powered applications running locally on PCs for improved latency and privacy, potentially lowering reliance on cloud-based solutions.

Intel is not alone in its ambitions. Rival AMD is also keen on powering the AI PC movement with its recently released Ryzen 7040 CPUs for laptops, and Qualcomm wants to make a bigger show of its built-in AI acceleration capabilities with the upcoming Snapdragon X chips. Apple’s homegrown M-series processors for Mac computers benefit too from on-chip AI acceleration.

But Intel believes it has substantial resources to enable AI PCs that can’t be matched by rivals.

“We believe this is going to unlock 100 million AI PCs over the next couple of years by 2025, and that’s a scale that we think our competition can’t reach,” Andrew Marsee, general manager of Intel Partner Alliance, told CRN in an interview.

If Intel and other companies succeed in a making a strong case for AI PCs, it could potentially create a new wave of refresh opportunities for channel partners in the commercial and consumer segments.

“This just unlocks new business opportunities, new business models that we probably haven’t even dreamt up yet,” Marsee said.

What Intel’s AI PC Acceleration Program Hopes To Achieve

The aim of the AI PC Acceleration Program is to connect ISVs as well as independent hardware vendors with Intel resources. These include co-engineering and design resources, AI toolchains, technical expertise, hardware and co-marketing opportunities.

The Santa Clara, Calif.-based chipmaker hopes these resources will help ecosystem partners “take full advantage” of the upcoming Core Ultra processors—which will feature an NPU for low-power AI workloads—and “corresponding hardware.”

In turn, Intel thinks this will “maximize AI and machine learning application performance, accelerate new use cases and connect the wider PC industry to the solutions emerging in the AI PC ecosystem.”

While Intel hopes ISVs in the program will use the company’s own tools, such as OpenVINO—which optimizes AI inference for a variety of processors—it’s not putting any requirements on what kind of tools or frameworks developers want to use for their applications.

“It’s more about making sure these applications are enabled and running and work well to raise all boats,” Robert Hallock, a former AMD staffer who is now senior director of client technology and performance marketing at Intel, told CRN. “And yes, of course, we get a little benefit because we help them do it, but we recognize that the need is bigger than us.”

Hallock said Intel also won’t block ISVs from optimizing their applications for chip architectures from competitors. However, he said that the NPU in the upcoming Core Ultra processors is more flexible than rival chips with the type of computations it can handle.

“The reality is, ours is just the most flexible, and that may result in [an application] running best on Intel naturally,” he said.

When new AI-powered applications and features are ready for the limelight, Intel will help ISVs connect with partners of all types for new sales opportunities.

“We will activate our channels for AI PCs, from consumer through business,” Marsee said.