Inside a sprawling lab at Google headquarters in Mountain View, California, hundreds of server racks hum across several aisles, performing tasks far less ubiquitous than running the world’s dominant search engine or executing workloads for Google Cloud’s millions of customers.
Instead, they’re running tests on Google’s own microchips, called Tensor Processing Units, or TPUs.
Originally trained for internal workloads, Google’s TPUs have been available to cloud customers since 2018. In July, Apple revealed it uses TPUs to train AI models underpinning Apple Intelligence. Google also relies on TPUs to train and run its Gemini chatbot.
“The world sort of has this fundamental belief that all AI, large language models, are being trained on Nvidia, and of course Nvidia has the lion’s share of training volume. But Google took its own path here,” said Futurum Group CEO Daniel Newman. He’s been covering Google’s custom cloud chips since they launched in 2015.
Google was the first cloud provider to make custom AI chips. Three years later, Amazon Web Services announced its first cloud AI chip, Inferentia. Microsoft‘s first custom AI chip, Maia, wasn’t announced until the end of 2023.
But being first in AI chips hasn’t translated to a top spot in the overall rat race of generative AI. Google’s faced criticism for botched product releases, and Gemini came out more than a year after OpenAI’s ChatGPT.
Google Cloud, however, has gained momentum due in part to AI offerings. Google parent company Alphabet reported cloud revenue rose 29% in the most recent quarter, surpassing $10 billion in quarterly revenues for the first time.
“The AI cloud era has completely reordered the way companies are seen, and this silicon differentiation, the TPU itself, may be one of the biggest reasons that Google went from the third cloud to being seen truly on parity, and in some eyes, maybe even ahead of the other two clouds for its AI prowess,” Newman said.
‘A simple but powerful thought experiment’
In July, CNBC got the first on-camera tour of Google’s chip lab and sat down with the head of custom cloud chips, Amin Vahdat. He was already at Google when it first toyed with the idea of making chips in 2014.
Amin Vahdat, VP of Machine Learning, Systems and Cloud AI at Google, holds up TPU Version 4 at Google headquarters in Mountain View, California, on July 23, 2024.
Marc Ganley
“It all started with a simple but powerful thought experiment,” Vahdat said. “A number of leads at the company asked the question: What would happen if Google users wanted to interact with Google via voice for just 30 seconds a day? And how much compute power would we need to support our users?”
“We realized that we could build custom hardware, not general purpose hardware, but custom hardware — Tensor Processing Units in this case — to support that much, much more efficiently. In fact, a factor of 100 more efficiently than it would have been otherwise,” Vahdat said.
Google data centers still rely on general-purpose central processing units, or CPUs, and Nvidia’s graphics processing units, or GPUs. Google’s TPUs are a different type of chip called an application-specific integrated circuit, or ASIC, which are custom-built for specific purposes. The TPU is focused on AI. Google makes another ASIC focused on video called a Video Coding Unit.
The TPU, however, is what set Google apart. It was the first of its kind when it launched in 2015. Google TPUs still dominate among custom cloud AI accelerators, with 58% of the market share, according to The Futurum Group.
Google coined the term based on the algebraic term “tensor,” referring to the large-scale matrix multiplications that happen rapidly for advanced AI applications.
With the second TPU release in 2018, Google expanded the focus from inference to training and made them available for its cloud customers to run workloads, alongside market-leading chips such as Nvidia’s GPUs.
“If you’re using GPUs, they’re more programmable, they’re more flexible. But they’ve been in tight supply,” said Stacy Rasgon, senior analyst covering semiconductors at Bernstein Research.
The AI boom has sent Nvidia’s stock through the roof, catapulting the chipmaker to a $3 trillion market cap in June, surpassing Alphabet and jockeying with Apple and Microsoft for position as the world’s most valuable public company.
“Being candid, these specialty AI accelerators aren’t nearly as flexible or as powerful as Nvidia’s platform, and that is what the market is also waiting to see: Can anyone play in that space?” Newman said.
Now that we know Apple’s using Google’s TPUs to train its AI models, the real test will come as those full AI features roll out on iPhones and Macs next year.
Broadcom and TSMC
It’s no small feat to develop alternatives to Nvidia’s AI engines. Google’s sixth generation TPU, called Trillium, is set to come out later this year.
Google showed CNBC the sixth version of its TPU, Trillium, in Mountain View, California, on July 23, 2024. Trillium is set to come out later in 2024.
Marc Ganley
“It’s expensive. You need a lot of scale,” Rasgon said. “And so it’s not something that everybody can do. But these hyperscalers, they’ve got the scale and the money and the resources to go down that path.”
The process is so complex and costly that even the hyperscalers can’t do it alone. Since the first TPU, Google’s partnered with Broadcom, a chip developer that also helps Meta design its AI chips. Broadcom says it’s spent more than $3 billion to make these partnerships happen.
“AI chips — they’re very complex. There’s lots of things on there. So Google brings the compute,” Rasgon said. “Broadcom does all the peripheral stuff. They do the I/O and the SerDes, all of the different pieces that go around that compute. They also do the packaging.”
Then the final design is sent off for manufacturing at a fabrication plant, or fab — primarily those owned by the world’s largest chipmaker, Taiwan Semiconductor Manufacturing Company, which makes 92% of the world’s most advanced semiconductors.
When asked if Google has any safeguards in place should the worst happen in the geopolitical sphere between China and Taiwan, Vahdat said, “It’s certainly something that we prepare for and we think about as well, but we’re hopeful that actually it’s not something that we’re going to have to trigger.”
Protecting against those risks is the primary reason the White House is handing out $52 billion in CHIPS Act funding to companies building fabs in the U.S. — with the biggest portions going to Intel, TSMC, and Samsung to date.
Processors and power
Google showed CNBC its new Axion CPU,
Marc Ganley
“Now we’re able to bring in that last piece of the puzzle, the CPU,” Vahdat said. “And so a lot of our internal services, whether it’s BigQuery, whether it’s Spanner, YouTube advertising and more are running on Axion.”
Google is late to the CPU game. Amazon launched its Graviton processor in 2018. Alibaba launched its server chip in 2021. Microsoft announced its CPU in November.
When asked why Google didn’t make a CPU sooner, Vahdat said, “Our focus has been on where we can deliver the most value for our customers, and there it has been starting with the TPU, our video coding units, our networking. We really thought that the time was now.”
All these processors from non-chipmakers, including Google’s, are made possible by Arm chip architecture — a more customizable, power-efficient alternative that’s gaining traction over the traditional x86 model from Intel and AMD. Power efficiency is crucial because, by 2027, AI servers are projected to use up as much power every year as a country like Argentina. Google’s latest environmental report showed emissions rose nearly 50% from 2019 to 2023 partly due to data center growth for powering AI.
“Without having the efficiency of these chips, the numbers could have wound up in a very different place,” Vahdat said. “We remain committed to actually driving these numbers in terms of carbon emissions from our infrastructure, 24/7, driving it toward zero.”
It takes a massive amount of water to cool the servers that train and run AI. That’s why Google’s third-generation TPU started using direct-to-chip cooling, which uses far less water. That’s also how Nvidia’s cooling its latest Blackwell GPUs.
Despite challenges, from geopolitics to power and water, Google is committed to its generative AI tools and making its own chips.
“I’ve never seen anything like this and no sign of it slowing down quite yet,” Vahdat said. “And hardware is going to play a really important part there.”
A former EU commissioner has hit back after receiving a U.S. visa ban for alleged censorship.
The Trump administration imposed visa bans on Thierry Breton, a former European Union commissioner behind the Digital Services Act (DSA), and four anti-disinformation campaigners, accusing them of censoring U.S. social media platforms.
“The State Department is taking decisive action against five individuals who have led organized efforts to coerce American platforms to censor, demonetize, and suppress American viewpoints they oppose,” Secretary of State Marco Rubio said in a statement.
He added that “these radical activists and weaponized NGOs have advanced censorship crackdowns by foreign states—in each case targeting American speakers and American companies.”
As such, their entry to the U.S. has “potentially serious adverse foreign policy consequences,” he said.
“Based on these determinations, the Department has taken steps to impose visa restrictions on agents of the global censorship-industrial complex who, as a result, will be generally barred from entering the United States.”
Breton, who served as EU commissioner between 2019 and 2024, wrote on X: “As a reminder: 90% of the European Parliament — our democratically elected body — and all 27 Member States unanimously voted the DSA.”
“To our American friends: “Censorship isn’t where you think it is.””
Rubio did not identify who his department had taken action against, however Under Secretary for Public Diplomacy Sarah Rogers later did so on X.
Josephine Ballon, the co-leader of HateAid who serves on Germany’s Advisory Council of the Digital Services, was among those working on anti-disinformation campaigns to receive sanctions. Her co-leader Anna-Lena von Hodenberg was also affected. CNBC has reached out to Ballon and Von Hodenberg for comment.
The bans are part of efforts to enforce what Rogers refers to as a “red line” for the U.S. and the “extraterritorial censorship of Americans.”
In an interview with GB news on Dec. 4, Rogers took aim at the U.K.’s Online Safety Act (OSA), saying the law was being applied extraterritorially, accounting for U.S. citizens’ speech about U.S. politics on U.S.-based platforms.
Europe’s DSA and the U.K.’s OSA are among only a handful of pieces of legislation designed to keep the power of Big Tech in check and improve safety for children online.
The DSA forces tech giants like Google and Meta to police illegal content more aggressively, or face hefty fines, while the OSA law requires age verification on adult sites and a number of other platforms.
A Waymo car is halted on the road amid a power outage in San Francisco, California, U.S., December 20, 2025, in this screengrab obtained from a social media video.d
Reuters
Three days after a blackout in San Francisco caused Waymo to pause its driverless car service, the Alphabet-owned company said it’s updating its fleet so its vehicles are better prepared to respond during future outages.
“We’ve always focused on developing the Waymo Driver for the world as it is, including when infrastructure fails,” the company said in a blog post late Tuesday.
Power outages began early afternoon on Saturday in San Francisco and peaked roughly two hours later, affecting about 130,000 customers, according to Pacific Gas and Electric. As of Sunday morning, about 21,000 customers remained without power. PG&E said a fire at a substation resulted in “significant and extensive” damage.
With stoplights and traffic signals not functioning, the city was hit with widespread gridlock. Videos shared on social media appeared to show multiple Waymo vehicles stalled in traffic in various neighborhoods.
“We directed our fleet to pull over and park appropriately so we could return vehicles to our depots in waves,” Waymo said in Tuesday’s blog post. “This ensured we did not further add to the congestion or obstruct emergency vehicles during the peak of the recovery effort.”
San Francisco Mayor Daniel Lurie said in an update on X Saturday evening that police officers, fire crews, parking control officers and city ambassadors were deployed across affected neighborhoods.
Waymo said that it’s analyzing the event, and is taking three “immediate steps.”
The first involves “fleet-wide updates” to give vehicles “more context about regional outages,” so cars can take more decisive actions at intersections. The company said it’s also improving its “emergency response protocols,” and is coordinating with Mayor Lurie’s team in San Francisco to better collaborate in emergency preparedness. Finally, Waymo said it’s updating its first responder training “as we discover learnings from this and other widespread events.”
In addition to the Bay Area, Waymo currently serves paid rides to the public in and around Austin, Texas, Phoenix, Atlanta and Los Angeles. The company recently crossed an estimated 450,000 weekly paid rides, and said in December it had served 14 million trips in 2025, putting it on pace to end the year at more than 20 million trips total since launching in 2020.
“Backed by 100M+ miles of fully autonomous driving experience and a record of improving road safety, we are undaunted by the opportunity to challenge the status quo of our roads, and we’re proud to continue serving San Franciscan residents and visitors,” the company said in Tuesday’s blog.
— CNBC’s Lora Kolodny and Jennifer Elias contributed to this report.
Direxion signage at the New York Stock Exchange (NYSE) in New York, US, on Monday, Dec. 22, 2025. The holiday-shortened week started with gains in stocks amid a broad advance that saw a continuation of the bullish momentum on Wall Street.
Michael Nagle | Bloomberg | Getty Images
Motive, a company with software for managing corporate trucks and drivers, on Tuesday filed for an initial public offering on the New York Stock Exchange under the symbol “MTVE.”
The paperwork puts Motive among a fast-growing group of tech companies looking to go public in 2026. Anthropic, OpenAI and SpaceX have all reportedly considered making their shares widely available for trading next year.
Motive is smaller, reporting a $62.7 million net loss on $115.8 million in revenue in the third quarter. The loss widened from $41.3 million in the same quarter of 2024, while revenue grew about 23% year over year. The company had almost 100,000 clients at the end of September.
Ryan Johns, Obaid Khan and Shoaib Makani started Motive in 2013, originally under the name Keep Truckin. Makani, the CEO, is Khan’s brother-in-law.
Investors include Alphabet’s GV, Base10 Partners, Greenoaks, Index Ventures, Kleiner Perkins and Scale Venture Partners.
Motive’s AI Dashcam device for detecting unsafe driving “has prevented 170,000 collisions and saved 1,500 lives on our roads,” Makani wrote in a letter to investors. Most revenue comes from subscriptions, although Motive does sell replacement hardware and professional services.
The San Francisco company changed its name to Motive in 2022, and as of Sept. 30, it employed 4,508 people. Motive employs 400 full-time data annotators who apply labels that are meant to enhance artificial intelligence models.
Motive has ongoing patent-infringement litigation with competitor Samsara, which went public in 2021 and today has a $22 billion market capitalization.