Meta and Broadcom just dropped a deal that feels like it’s trying to quiet a lot of investor noise at once: the partnership is getting extended through 2029, and Meta is promising an initial deployment of 1 gigawatt of its Training and Inference Accelerators.
The announcement also comes with a board-level detail that can’t really be ignored. Meta said Broadcom CEO Hock Tan told Meta last week that he has decided not to stand for reelection to Meta’s board, according to a filing. Tan joined Meta’s board in 2024. Even if you’re not tracking corporate governance day-to-day, you can usually feel when a major tech relationship shifts from “business as usual” to something more deliberate.
Meta’s commitment is tied to what the companies are calling a custom in-house chip roadmap. The MTIA chips will be the first AI silicon to use a 2 nanometer process, Broadcom said in its own statement. Over time, the deal is set to see Meta deploying multiple gigawatts of chips based on Broadcom technology—so not just a token pilot, but a scaling plan that targets the kind of compute demand that keeps climbing as AI data centers expand.
“Meta is partnering with Broadcom across chip design, packaging, and networking to build out the massive computing foundation we need to deliver personal superintelligence to billions of people,” Meta’s co-founder and CEO, Mark Zuckerberg, was quoted as saying in the statement. In the background of all that big ambition, the stock-market reaction was a little calmer than the language: Broadcom shares rose 3% in extended trading after the announcement, while Meta stock was flat.
The part that really grabbed attention, though, was the pushback embedded in the statements. Tan said on Broadcom’s March earnings call: “Now, contrary to recent analyst reports, Meta’s custom accelerator, MTIA roadmap is alive and well. We’re shipping now and, in fact, for the next generation XPUs, we will scale to multiple gigawatts in 2027 and beyond,” Tan said. That’s a direct response to chatter—like someone stepping onto a stage and saying, no, the show is still happening.
Meta has been moving fast on its own chip efforts, unveiling four new versions of its in-house MTIA chips in March. It first unveiled the custom silicon in 2023, following on the heels of similar chip programs at Google and Amazon. The broader reason hyperscalers are obsessed with alternatives is pretty plain: GPUs—especially the costly, constrained ones from Nvidia and AMD—can be expensive and hard to scale smoothly. So the industry keeps circling back to application-specific integrated circuits, or ASICs, which are smaller and cheaper than general-purpose AI workhorse GPUs, but limited to narrower tasks. And in Meta’s case, MTIA is used entirely for internal purposes, not sold as a cloud option.
This latest deal also lands shortly after Broadcom announced a long-term agreement with Google for producing its TPUs, and said Anthropic would access 3.5 gigawatts worth of the in-house Google chip. Broadcom shares are up 10% so far in 2026, while the S&P 500 index has gained about 2% over the same period. Meta’s board changes continue in parallel, too: Tracey Travis will leave Meta’s board after taking a board seat in 2020, Meta said.
Meta’s spending tempo doesn’t slow down either. The company has made a flurry of deals since committing in January to spending up to $135 billion on AI this year. Those include commitments to deploy up to 6 gigawatts of AMD GPUs, millions of Nvidia chips and new custom chips made by chip architecture firm Arm Holdings. And yes, Meta has plans for 31 data centers, including 27 in the U.S. Somewhere in a server room, the fans are already doing their usual steady roar—nothing cinematic, just constant work—while the corporate timeline keeps marching forward. Then again, maybe the real story is that the compute race is so tight now that even the smallest roadmap wobble can turn into a headline.
Western United Cup returns to Geelong, eyes July Tarneit event
Super Typhoon Sinlaku Batters US Mariana Islands, Guam
Kerala Blasters vs NorthEast United: A relegation test on Vishu