Google just dropped a number that's reshaping the AI infrastructure landscape - the company's planning to spend up to $185 billion on capital expenditures this year, nearly double last year's outlay. The announcement sent Broadcom shares climbing 6% in after-hours trading Wednesday, while Nvidia gained 2%. The market reaction reveals something crucial about the AI hardware race: it's not winner-takes-all, and Google's custom chip strategy is making Broadcom a massive beneficiary.
Google just rewrote the playbook on AI infrastructure spending, and Wall Street's scrambling to figure out who wins. The company's bombshell earnings announcement Wednesday revealed plans to spend as much as $185 billion on capital expenditures this year - a staggering near-doubling from 2025 levels that's reshaping the semiconductor landscape in real time.
Broadcom shares surged 6% in extended trading, while Nvidia climbed 2%. But the real story isn't just the rising tide lifting all boats. It's about how Google's massive bet on custom silicon is creating a parallel universe to Nvidia's GPU dominance, with Broadcom sitting at the center of it.
"That is an incredible number. We are laughing because that number is so good for the Google cohort," Ben Reitzes, Melius Research head of technology research, told CNBC's Closing Bell Overtime following the release. The "Google cohort" he's referring to includes a tight circle of suppliers building the infrastructure for Google's AI ambitions.
Here's what most people miss: much of Google's cutting-edge AI work doesn't run on industry-standard Nvidia chips at all. The company's state-of-the-art Gemini 3 model was trained on tensor processing units, Google's proprietary chips designed specifically for machine learning workloads. And Broadcom is the critical partner helping Google design and manufacture these TPUs.
Broadcom has quietly built a burgeoning custom chip business focused on application-specific integrated circuits, or ASICs, which some experts believe can be more efficient than general-purpose GPUs for certain AI tasks. In December, Broadcom revealed it would sell Google's TPU Ironwood rack systems to Anthropic, the AI lab behind Claude. The move essentially turned Google's internal infrastructure into a product line.
The custom chip game only makes economic sense at massive scale - what the industry calls "hyperscaler" territory. Broadcom refers to the custom AI accelerators it's developing for five separate customers as "XPUs," though it's only publicly named Google and Anthropic. Industry watchers suspect Microsoft, Amazon, and Meta are also working with Broadcom on their own custom silicon projects.
These tech giants need partners like Broadcom because designing chips is only half the battle. Hyperscalers require specialized intellectual property and manufacturing expertise that semiconductor companies have spent decades building. Broadcom adds the critical IP blocks and coordinates with foundries to actually produce the chips at scale.
But Google isn't putting all its chips in one basket, so to speak. The company also relies on Nvidia's hardware for portions of its AI infrastructure, which explains why Nvidia shares also rose in after-hours trading despite Google's TPU focus.
"It's probably good for Nvidia too because they are going to spread the love, not just their own TPU but also to Nvidia," Reitzes said. The comment captures the emerging reality of AI infrastructure: even companies building custom chips need Nvidia's ecosystem for flexibility and certain workloads.
The $185 billion figure represents a seismic shift in capital allocation across the tech industry. Google is part of a broader trend where Microsoft, Amazon, and Meta are all dramatically increasing data center spending to power AI development. But Google's approach - betting heavily on custom silicon while maintaining Nvidia partnerships - offers a blueprint for how hyperscalers can avoid total dependence on a single chip supplier.
For Broadcom, the strategy is paying off spectacularly. The company's custom chip division has transformed from a niche business into a growth engine, with Google's massive capex providing years of runway. The TPU partnership alone represents billions in potential revenue, and Broadcom's work with other hyperscalers multiplies that opportunity.
The market's reacting to more than just one quarter's guidance. Google's spending commitment signals a multi-year infrastructure build-out that will require continuous chip design iterations, manufacturing capacity, and support. Broadcom's locked into that cycle as Google's design partner, while Nvidia benefits from Google's need for GPU capacity alongside its custom chips.
What's emerging is a two-tier AI chip market: Nvidia's GPUs as the flexible, general-purpose standard, and custom ASICs from partners like Broadcom for hyperscalers optimizing specific workloads at massive scale. Google's spending proves there's room for both, and the dollars flowing into AI infrastructure are large enough to support multiple winners.
Google's nearly $185 billion AI infrastructure commitment isn't just a win for Broadcom and Nvidia - it's a validation of the dual-track approach to AI chips. While Nvidia's GPUs remain the industry standard, custom silicon partnerships are proving economically viable at hyperscale, and the capex dollars are large enough to support both ecosystems. For investors, the takeaway is clear: the AI infrastructure build-out is creating multiple winners, and companies like Broadcom that enable custom chip strategies are positioned to capture massive revenue streams alongside GPU giants. Watch how other hyperscalers respond to Google's spending levels - if Microsoft, Amazon, and Meta follow suit with similar capex increases, we're looking at a multi-year semiconductor supercycle that extends far beyond Nvidia's dominance.
