Broadcom claims a vast majority of its largest customers have signed up for VMware’s flagship Cloud Foundation (VCF) platform, further solidifying its controversial move to revamp VMware’s operating structure and laying the foundation for further growth tied to surging artificial intelligence (AI) adoption.

Broadcom CEO Hock Tan told investors during the vendor’s latest earnings call that 70% of its largest 10,000 customers have now adopted the VCF platform. That was considerable progress from the 4,500 large customers Tan said had signed on for the VCF platform at the end of the previous quarter.

Tan noted during Broadcom’s previous earnings call that the vendor had booked 21 million total CPU cores during compared to 19 million cores posted in the preceding quarter. More significantly, Tan said that 70% of those new booked cores were on its VCF platform, “virtualizing the entire data center.”

That VMware adoption growth comes as the broader ecosystem continues to work through Broadcom’s changes.

Analysts and rivals have over the past year been pointing to an expected – and real – outflow of VMware customers stung by Broadcom’s pricing and license changes. However, more recent commentary has noted Broadcom’s success in retaining VMware’s largest customers.

Prashanth Shenoy, VP of cloud platform, infrastructure, and solutions marketing at Broadcom, recently explained to SDxCentral that the vendor is focused this year on “acceleration and adoption” of its VMware business. That focus is expected to ride momentum coming out of a hectic 2024, which saw Broadcom overhaul VMware’s platforms and licensing models.

“Last year was a year of cleaning up, simplifying our route to market, our offer, our product, giving them the direction, the confidence, and building the product,” Shenoy said of Broadcom’s updates to the VCF platform. “This year is all about making sure the product gets deployed, consumed, all of the advanced services work on that, and our partners and our route to market are fully operationalized to go deliver on that.”

That VMware operationalization could also feed Broadcom’s overall AI business backed by the VMware Private AI Foundation platform launched in 2023.

“As large enterprises adopt AI, they have to run their AI workloads on their [on-premises] data centers, which will include both GPU servers as well as traditional CPUs,” Tan said during the latest earnings call about that VMware Private AI Foundation platform. “And just as VCF virtualizes these traditional data centers using CPUs, VCF will also virtualize GPUs on a common platform and enable enterprises to import AI models to run their own data [on-premises].”

Tan added that 39 enterprise customers were currently using the platform.

Too busy for Intel? With VMware now seemingly well integrated, Tan was asked about rumors surrounding Broadcom’s future acquisition appetite.

“I'm too busy,” Tan retorted. “We're too busy doing AI and VMware at this point. We're not thinking of it at this point.”

Broadcom has reportedly been in talks to acquire a portion of beleaguered chip giant Intel. Those reports suggested Broadcom would acquire Intel’s chip-design business and that Taiwan Semiconductor Manufacturing Co. (TSMC) might purchase Intel’s production facilities.

Tan during Broadcom’s previous earnings call stated the vendor did not have enough cash on hand for a “big” acquisition but added that it was always open to opportunities.

Tan cautioned that Broadcom’s acquisition approach was in “buying big enough companies” that its current cash on hand was “not adequate” to move down that road. However, he did add that did not mean it was not possible.

“We are open. Still open, always open, because that’s been a core part of our strategy, business model of this company for the last 10 years, which is we’re always interested in adding to our portfolio, very good franchise assets, be they in semiconductors or be they in infrastructure software, as long as they meet the criteria, the fairly demanding criteria we look for, we will always be open to acquiring assets and adding into a portfolio,” Tan said.

Broadcom is already deeply embedded in the chip ecosystem and has touted its position within the growing AI opportunity. That is also an opportunity Intel is targeting.

Broadcom’s overall AI-related revenues surged 220% year over year to $12.2 billion during its fourth fiscal quarter of 2024, which Tan tied to its custom AI accelerators and its networking business. These AI components accounted for 41% of Broadcom’s overall semiconductor revenue during that quarter and powered overall semiconductor revenue to a record $30.1 billion for the full fiscal year.

Tan stated that AI was set to become Broadcom’s dominant semiconductor driver over the next several years. This will be driven by Broadcom’s hyperscaler partners that are themselves driving significant investments into their AI-focused infrastructure and are relying heavily on Broadcom’s XPU architecture.

“We currently have three hyperscale customers who have developed their own multigenerational AI XPU roadmap to be deployed at varying rates over the next three years,” Tan said during Broadcom’s earnings call. “In 2027, we believe each of them plans to deploy one million XPU clusters across a single fabric.”

Tan added that this will feed into a total addressable market of up to $90 billion by 2027.

“The reality going forward for this company is that the AI semiconductor business will rapidly outgrow the non-AI semiconductor business,” Tan said.

Equity research firm William Blair noted in a report that Broadcom supplies custom chips to Google, Meta, and Bytedance, and also noted that Broadcom hinted at a pair of new hyperscale customers it was working through a validation process with that the analyst firm suspects are OpenAI and Apple.

“Altogether, Broadcom’s commentary on its AI opportunity with customers was a positive reinforcement for our AI thesis, highlighting that despite some ongoing investor concerns about scaling laws and pre-/post-training compute needs, the largest hyperscalers remain steadfast in their plans to continue scaling out their AI clusters over the next few years,” William Blair noted in its report.