In an interview with CRN, Intel U.S. Channel Chief Michael Green admits that the rollout of the recently launched Gaudi 3 is a ‘slow process’ as the company scales up OEM support for the accelerator chip. ‘It’s a brand-new product that we’re ramping. We have visions of making this a channel-available product in 2025,’ he says.

Intel U.S. Channel Chief Michael Green said channel partners will play a “massive role” in the rollout of the semiconductor giant’s Gaudi 3 accelerator chip when it becomes a “channel-available product in 2025.”

Promised by Intel to enable more cost-effective AI systems than those using Nvidia GPUs, Gaudi 3 launched last month in servers from Dell Technologies and Supermicro. Hewlett Packard Enterprise is expected to follow with its own Gaudi 3 system in December. Intel has not yet said when servers from other vendors, including Lenovo, will arrive.

[Related: Analysis: Intel’s AI Chip Efforts Stall As AMD Gets A Boost Against Nvidia]

In a recent interview with CRN, Michael Green, who became the head of Intel’s new North America partner scale group this month, admitted that the rollout of Gaudi 3 is a “slow process” as the company scales up OEM support for a product line that was previously supported by only one OEM, Supermicro, with the predecessor chip. As a result, Intel’s channel partners shouldn’t expect to sell Gaudi 3 systems until next year.

“We’re already working, obviously, with the Dells and the Supermicros, really the Dells and the HPEs of the world, on making sure that it gets into partners’ hands. But let’s be honest: It’s a slow process, right? It’s a brand-new product that we’re ramping. We have visions of making this a channel-available product in 2025,” he said.

“The channel plays a massive role in how do we take Gaudi 3 out to the masses,” he added. “But if you go from Gaudi 2 where we had one OEM partner to now having four, it’s massive progress. So then how do we take that and make it more applicable to the channel partners? In our strategy, it will just take a little bit of time.”

Beyond Intel’s pitch for Gaudi 3 delivering a “price performance advantage” in AI systems, the chipmaker has said the chip will find favor with enterprises because of Intel’s open ecosystem approach across hardware infrastructure, software platforms and applications.

For software, this open approach means that partners and customers can choose from a variety of tools from different vendors to address every software need for an AI system, according to Intel. A key aspect of this approach is the Open Platform for Enterprise AI, a Linux Foundation group that is meant to serve as the foundation for microservices from different vendors that can be used for such systems.

Green said Intel will play a “key role” in helping channel partners connect with ISVs to build out the necessary software stack for Gaudi 3 systems.

“We need to be the leading voice and the leading aggregator of all these partners in ISVs, because collectively, this is where we’ll actually see improvements on what’s available to end users, what they can do from a cost-optimization standpoint,” he said.

Partners Welcome Gaudi 3 Channel Plan—With Some Skepticism

Solution provider executives who spoke to CRN said Intel is on point with its plan to make a big push for Gaudi 3 in the channel, especially when demand for AI infrastructure and the need for alternatives to Nvidia GPUs remain high. But they weren’t in agreement on how much of an opportunity Gaudi 3 could represent for Intel.

“At the end of the day, the channel is an amazing vehicle to get things not only launched, but embedded and driven,” said Bob Venero, CEO of Fort Lauderdale, Fla.-based Future Tech Enterprise, No. 76 on CRN’s 2024 Solution Provider 500 list. “If history shows itself, we’ll be able to make a big impact for them with the product.”

Venero thinks Intel’s Gaudi 3 chip could make inroads with enterprises if the semiconductor giant makes an emphasis on how it could enable AI data centers with lower operational costs, such as energy, than those powered by Nvidia GPUs.

“You look at companies that are trying to do AI farms and stuff, there’s just not enough power or cooling to do it, and that’s a big thing. So if I’m Intel, I’m driving that messaging on not just price but operational price,” he said.

The solution provider CEO said he thinks Intel could also benefit from its existing relationships in the enterprise world, but all of this rests on the chipmaker’s ability to develop a program that will enable partners to sell Gaudi 3 systems.

“You’re not getting fired by putting Intel in your environment, right? So I think they’ll have a lot of opportunity and growth, especially out of the gate once the partners are able to drive it and Intel helps build a program that allows them to drive it,” Venero said.

Alexey Stolyar, CTO of Northbrook, Ill.-based systems integrator International Computer Concepts, told CRN while he believes there is appetite for alternatives to AI systems powered by Nvidia systems, he’s not yet sure how big of an opportunity is represents.

While Stolyar said he is working on one Gaudi 3 project for a customer, customers are still mostly talking about Nvidia’s latest GPUs, including the forthcoming Blackwell chips and their associated platforms, like the GB200 NVL72 rack-scale system.

“It’s not proportional yet,” he said.

Stolyar said the difficulty with selling AI infrastructure to enterprises is that they are typically slower than startups and hyperscalers when it comes to adopting new technologies, especially when the return on investment still isn’t clear around AI.

“Most of them don’t know how to utilize it yet, don’t know how to build the systems, don’t know how to do anything with them yet, and there’s a pretty big learning curve. Everyone wants to, but they’re trying to figure out, what does that mean from a money perspective? How are they going to generate money from it?” he said.

Dominic Daninger, vice president of engineering at Burnsville, Minn.-based Nor-Tech, told CRN that Intel’s biggest challenge in getting the market to adopt Gaudi 3 is Nvidia’s incumbency, especially in light of how the rival has expanded its software offerings.

“It’ll be a struggle because Nvidia has dominated it for so long and got so much software infrastructure there that people have adopted,” he said.

One thing that could help Intel, according to Daninger, is the fact that Supermicro has started to make Gaudi 3 systems available through its JumpStart program for remote testing. But the executive added that his company hasn’t given much thought to selling Gaudi 3 systems yet because customers haven’t asked about them.

“We would if we saw any interest or demand,” he said.

A senior leader at a solution provider that is in the top 100 of CRN’s Solution Provider 500 said that the channel will be a critical route to market for Intel’s Gaudi 3 because that’s how most enterprises buy IT infrastructure, and they will represent a larger opportunity than the smaller group of companies making mass purchases today.

“That’s where I think the impetus is on them to get us enabled, because we’re going to be the ones moving the needle probably more so than they are,” said the executive, who asked to not be named to speak candidly.

But while the solution provider said Gaudi 3 could benefit from the need for more alternatives to Nvidia GPUs, he has observed that Intel’s other rival, AMD, has been much more active in engaging with partners lately.

“Their sellers are out with our sellers. They are trying to proactively win that affinity battle. And Intel’s just been a little too lax in that that respect,” said the executive, who added that Intel’s massive job cuts have been a continued source of concern.