As Intel Creates New AI Group, Data Center Division To ‘Refocus’ On CPUs: Memos

In memos to Intel employees that were seen by CRN, company leaders indicate that what has been known as the Data Center and AI Group is being split up, with the newly renamed Data Center Group refocusing on CPUs and Sachin Katti taking over responsibilities for data center accelerator chips, like GPUs, in his freshly configured AI and CTO organization.

With Intel’s new CTO and AI chief taking over responsibilities for the company’s AI road map and strategy, the business unit most recently known as the Data Center and AI Group will “refocus” on CPUs and take on an older name, representing a major shift in how the company organizes two of its most critical product areas.

These changes were detailed in memos sent last Thursday to Intel employees by Sachin Katti (pictured), the company’s new chief AI and technology officer, and Karin Eibschitz, the interim general manager of what is now being called the Data Center Group.

The memos, which were seen by CRN, were sent the same day Intel CEO Lip-Bu Tan told employees in his own letter that he is flattening the executive leadership team’s structure to cut down on “organizational complexity and bureaucratic processes” that he said “have been slowly suffocating the culture of innovation we need to win.”

Reuters first reported on Tan’s memo Thursday. The memos by Katti and Eibschitz, which have not been previously reported, provide more details about the restructuring of Intel’s data center and AI teams that are critical to the company’s future.

An Intel spokesperson did not provide a comment by press time. The company said in a statement last Friday in response to Reuters’ report: “We continue to focus on fostering a culture of innovation across the company that empowers our engineering teams to create great products and delight our customers.”

Tan, who started as Intel’s CEO five weeks ago, said in his memo that the company’s business units, including the Client Computing Group and Data Center Group, will now report directly to him. This is a reversal from the previous structure announced in early December, when company veteran Michelle Johnston Holthaus was named CEO of Intel Products, a group that consisted of Intel’s core chip design business units. Holthaus will remain in her role and see her responsibilities expanded, according to Tan’s memo.

The CEO also expanded the roles and responsibilities for Katti— who was already part of Tan’s executive leadership team as the head of Intel’s Network and Edge Group—to include chief AI and technology officer. This will put him in charge of Intel’s “overall AI strategy and AI product road map as well as Intel Labs and our relationships with the startup and developer ecosystems,” according to Tan.

Tan made the changes in his bid to turn around Intel, which, among other things, has failed to find traction in the data center accelerator chip space dominated by Nvidia—whose annual revenue has surpassed Intel for the past two years—and is facing a growing number of competitive threats in the CPU market for data centers and PCs.

An executive at a national systems integrator that partners with Intel told CRN that Tan’s move to split responsibilities for data center CPUs and data center GPUs between two executive leaders could help Intel’s ability to focus on both areas.

But the executive, who asked to not be named to speak candidly, said he has concerns about how the arrangement could impact Intel’s ability to position CPUs against accelerator chips like GPUs for a wide range of AI solutions, especially since not every AI solution, such as computer vision, needs GPUs and could work well with just CPUs.

“If the [business units] can get on the same page and say, ‘This is what we’re taking to the market as far as where to use the CPU or GPU, and they work together, great. If not, they’re going to be just fighting and saying, ‘Yes, you need a GPU, or yes, you need a CPU,’ regardless of the application,” he said.

Intel Leaders Signal Split-Up Of Data Center And AI Group

In their respective memos to Intel employees, Eibschitz and Katti indicated that what has been known as the Data Center and AI Group since 2021 is being split up, with Eibschitz’s division continuing to handle data center CPUs and Katti taking over data center accelerator chips such as GPUs in his newly configured AI and CTO organization.

Eibschitz, who took over the division as its interim leader from Justin Hotard in February, referred to her business unit as the Data Center Group, a name that hasn’t been used for the division since 2019, before Intel restructured the group later that year and again under former CEO Pat Gelsinger in 2021 with the Data Center and AI Group name.

“In the Data Center Group, we will refocus on the fundamentals: creating exceptional products and delighting our customers,” wrote Eibschitz, who added that the group will “stay focused on re-energizing our CPU business.”

This focus will encompass both “broad-market Xeon-based solutions as well as custom x86-based products,” according to the Data Center Group leader.

“Our charter encompasses defining business strategy and road maps, product management, developing ecosystems and business opportunities, and strengthening [profits],” she wrote.

Eibschitz said the Data Center Group will collaborate with new horizontal engineering teams—reporting directly to Tan and led by long-time technical leaders Rob Bruckner, Mike Hurley and Lisa Peace—as well as Katti’s new CTO and AI organization.

“Thanks to all of you, we are rebuilding trust with our customers, delivering more competitive products and simplifying our road map,” she wrote. “However, we must continue to push forward with urgency to win in the data center market.”

Katti Says His Group Is At ‘Center’ Of Intel’s Future

In his memo to employees, Katti said he will “lead the strategy, definition and execution for our data center accelerator portfolio as well as product positioning and customer engagements” in his new role as chief technology and AI officer.

The executive, who has been at Intel for more than three years, said his group has absorbed Saurabh Kulkarni, vice president of AI systems design, and the AI systems and GPU product management team. This team was previously a part of the Data Center and AI Group, as Eibschitz noted in her memo.

Katti said the CTO and AI organization will also take in Anil Rao and the systems architecture and engineering team as well as what is called the Intel Cloud Services team. The latter team was most recently led by Markus Flierl, who launched the Intel Tiber AI Cloud service last year and “has decided to leave Intel to pursue external opportunities,” according to Katti. Katti said he plans to name Flierl’s successor.

Other teams joining Katti’s organization include Melissa Evers and the software ecosystem enablement team; Rich Uhlig and the Intel Labs team; and Hannah Kirby and the developer relations execution team, according to Katti’s memo. Some of these teams, such as Intel Labs and Intel Cloud Services, previously sat under Intel CTO Greg Lavender, who is being succeeded by Katti and is retiring, Tan said in his memo.

“These are important steps toward refining our AI strategy and attacking opportunities to gain greater relevance in the AI market,” Katti wrote. “This team is at the center of building a new Intel—one that is posed to lead, innovate and make a lasting impact. Let’s make it happen”

Katti didn’t discuss any specific plans for Intel’s AI strategy and road map in his memo. Prior to Tan becoming Intel’s CEO, the company in January canceled its Falcon Shores accelerator chip that was due by the end of this year and said it would focus on a “rack-scale solution” based on a successor chip called Jaguar Shores.

Katti Vows To ‘Evolve’ Network And Edge Group’s Approach

In his memo, Katti only dedicated one sentence to the Network and Edge Group—shortened as NEX—which he has had led as its general manager since 2023.

“Across NEX, we will continue to adapt and evolve our approach to positioning the business for the future,” he wrote.

The Network and Edge Group had already gone through substantial changes last September. That’s when Gelsinger announced that Intel would move its edge business out of the Network and Edge Group and into the Client Computing Group.

Gelsinger also moved Intel’s integrated photonics solutions out of the Network and Edge Group and into what was known as the Data Center and AI Group at the time.

This left the Network and Edge Group to sharpen its focus on networking and telecommunications, the now-former CEO said at the time.

Tan Says More Organizational Changes Will Come

In Tan’s memo to Intel employees, the CEO said he has asked each member of his executive team to “identify opportunities” for further organizational changes in support of his goals to reduce bureaucracy and speed up decision-making.

“As I said last month, Intel is going to act like a disruptor again by giving engineers the freedom to innovate,” he wrote in the memo, which was also seen by CRN.

“We will simplify the way we work so that it’s easier to get work done. And we will empower smaller, more focused teams to create great products and delight our customers,” Tan added.

Eibschitz and Katti indicated more changes are coming for their respective groups.

“Right now, we are implementing the highest-level [organization] changes while we continue to work through further refinements,” Eibschitz said of the Data Center Group.

“In the coming weeks, we will further refine our organizational structure to help us work together with greater speed and agility,” Katti wrote of his CTO and AI organization.