Anthropic secures 3.5 gigawatts of Google TPUs as revenue soars to $30 billion run rate
22 Sources
[1]
Anthropic ups compute deal with Google and Broadcom amid skyrocketing demand | TechCrunch
AI research lab Anthropic announced Monday that it signed a new agreement with Google and Broadcom for increased processing and compute capacity to power its Claude AI models. This reworking of its compute deals comes as demand for its AI models continues to soar. The deals would expand
[2]
Broadcom to supply Anthropic with 3.5 gigawatts of Google TPU capacity from 2027 -- Claude pioneer says its annual revenue run rate has passed $30 billion
Securities filing confirms multi-year supply agreement runs through 2031. Broadcom disclosed in a securities filing on Monday that it'll supply Anthropic with roughly 3.5 gigawatts of Google TPU capacity starting in 2027, and separately committed to designing and supplying future generations of
[3]
Anthropic reveals $30bn run rate, plan to use new Google TPU
Broadcom's building the silicon and is chuffed about that, but also notes Anthropic remains a risk Broadcom has announced that Google has asked it to build next-generation AI and datacenter networking chips, and that Anthropic plans to consume 3.5GW worth of the accelerators it delivers to the ads
[4]
Exclusive: Anthropic weighs building it own AI chips, sources say
SAN FRANCISCO, April 9 (Reuters) - Artificial intelligence lab Anthropic is exploring the possibility of designing its own chips, three sources said, as the company and its rivals respond to a shortage of AI chips needed to power and develop more advanced AI systems. The plans are in early stages
[5]
Anthropic strikes chips deal with Google and Broadcom
Anthropic will spend hundreds of billions of dollars on Google's chips and cloud services in a push to secure critical computing resources as surging demand for the company's tools pushes its annualised revenue to $30bn. The AI lab said on Monday it has committed to use "multiple gigawatts" of
[6]
Anthropic is exploring building its own AI chips
The plans are early-stage and Anthropic may still decide to only buy chips rather than design them. The exploration comes days after the company signed a long-term deal with Google and Broadcom for 3.5 gigawatts of TPU compute starting in 2027. A company spokesperson declined to comment. Anthropic
[7]
Anthropic signs biggest compute deal yet with Google and Broadcom as run rate hits $30bn | TNW
In short: Anthropic has agreed to access approximately 3.5 gigawatts of next-generation Google TPU compute capacity via Broadcom from 2027, its largest infrastructure commitment to date -- while simultaneously disclosing that its revenue run rate has surpassed $30bn, more than tripling from roughly
[8]
Anthropic Google Broadcom TPU deal 3.5 gigawatts and $30 billion revenue
Anthropic has signed an agreement with Google $GOOGL and Broadcom $AVGO for approximately 3.5 gigawatts of computing capacity built on Google's tensor processing units, with the capacity expected to come online starting in 2027, the company said. The deal expands an existing relationship. Broadcom
[9]
Anthropic reportedly mulls designing own chips amid shortage
Claude-creator Anthropic is considering designing its own chips, as advanced AI systems drive shortage, sources told Reuters. Anthropic continues to grab the headlines this week, as it fights the US administration in the courts, and the power of its unreleased Claude Mythos model strikes fear into
[10]
Anthropic taps Google and Broadcom for yet more AI chips as revenue run rate tops $30B - SiliconANGLE
Anthropic taps Google and Broadcom for yet more AI chips as revenue run rate tops $30B Anthropic PBC said today its annual revenue run rate has now exceeded $30 billion, up from just $9 billion at the end of last year, after confirming an expanded partnership with Google LLC and Broadcom Inc. to
[11]
Anthropic may develop its own AI chips
Anthropic is exploring the design of its own AI chips amid a shortage of necessary components for advanced AI development, according to Reuters. The company's plans are in early stages, and it may still opt to continue purchasing chips rather than developing them in-house. Anthropic has not yet
[12]
Anthropic, Google, Broadcom announce 3.5GW TPU deal
Revenue rate has surpassed $30bn at Anthropic, up from around $9bn at the end of 2025. A new expanded agreement will allow Anthropic to tap 3.5GW of Google's Tensor Processing Units (TPU) capacity from Broadcom. In a regulatory filing yesterday (6 April), Broadcom said that Anthropic's
[13]
Anthropic, Google, Broadcom strike 3.5GW TPU deal
Anthropic's revenue rate has exceeded $30 billion, a significant rise from approximately $9 billion at the end of 2025. This increase accompanies a newly expanded agreement allowing Anthropic to utilize 3.5 gigawatts (GW) of Google's Tensor Processing Units (TPU) capacity from Broadcom. A
[14]
Anthropic tops $30 billion run rate, seals deal with Broadcom - The Economic Times
Anthropic PBC said its revenue run rate has now topped $30 billion, up from $9 billion at the end of 2025, and confirmed plans to work with Broadcom and Google to power its burgeoning operations. The AI startup said that demand for its Claude services has accelerated this year, with more than
[15]
Anthropic Eyes Custom AI Chips Amid Global Shortage As Claude Demand Surges Past $30 Billion Run Rate: Re
Claude-parent Anthropic is reportedly in the preliminary stages of evaluating in-house chip development. Early-Stage Chip Ambitions The effort remains exploratory, with no formal commitment, finalized design, or dedicated team in place, Reuters reported on Thursday, citing people familiar with
[16]
Broadcom Signs Anthropic, Google TPU 'Groundbreaking' Deals To Drive AI Capacity
'Anthropic, beginning in 2027, will access through Broadcom approximately 3.5 gigawatts as part of the multiple gigawatts of next-generation TPU-based AI compute capacity committed by Anthropic,' says Broadcom in a government filing. Broadcom is making huge moves in the AI market as the $77
[17]
Anthropic Locks Multi-Gigawatt Compute Deal with Google and Broadcom Amid Surging AI Demand
The expansion represents one of Anthropic's largest compute commitments to date and underscores the intensifying competition among leading AI developers to secure long term high performance infrastructure at scale. Massive expansion of frontier AI compute capacity Under the new agreement,
[18]
Anthropic signs AI compute deal with Google and Broadcom
Revenue growth accelerates: The deal marks a significant expansion of Anthropic's infrastructure as demand for its Claude models increases. The company said its annualised revenue run rate (ARR) has crossed $30 billion in 2026, up from about $9 billion at the end of 2025. It also reported that more
[19]
Anthropic secures 3.5 GW AI Compute Deal with Google and Broadcom as Revenue Hits $30 Billion
The Artificial Intelligence (AI) race is entering a new phase of infrastructure growth, as Anthropic announced a long-term collaboration with Google and Broadcom, aiming to gain the next-generation AI compute capacity. The acquisition is timed with the annualized revenue run rate of Anthropic
[20]
Anthropic weighs building its own AI chips- Reuters By Investing.com
Investing.com-- Anthropic is considering designing its own artificial intelligence chips amid a shortage of processors needed to power advanced AI, Reuters reported on Thursday. The plans are still in early stages, and Anthropic may still decide to only buy AI chips over designing them, Reuters
[21]
Anthropic weighs building it own AI chips, sources say
SAN FRANCISCO, April 9 (Reuters) - Artificial intelligence lab Anthropic is exploring the possibility of designing its own chips, three sources said, as the company and its rivals respond to a shortage of AI chips needed to power and develop more advanced AI systems. The plans are in early stages
[22]
Anthropic may build its own AI chips amid rising demand and supply crunch
In-house chip plans are still early-stage with no final design or dedicated team yet Anthropic is again making headlines and this time it is for something new. As per the reports, the AI startup is reportedly considering designing its own AI chips. The company is reportedly aiming to reduce its
Share
Copy Link
Anthropic expanded its compute agreement with Google and Broadcom, securing 3.5 gigawatts of tensor processing unit capacity starting in 2027. The deal reflects explosive demand for Claude AI models, with the company's annualized revenue jumping from $9 billion to $30 billion in just three months. More than 1,000 business customers now spend over $1 million annually on Anthropic's services.
Anthropic Expands Google and Broadcom Deal to Secure Massive AI Compute Capacity
Anthropic announced Monday it has signed an expanded agreement with Google and Broadcom to secure approximately 3.5 gigawatts of additional AI compute capacity, marking the company's most significant infrastructure commitment to date
1
. The new capacity, which will come online in 2027, builds on the company's October 2025 deal for more than a gigawatt of compute and extends Anthropic's $50 billion commitment to invest in U.S. AI infrastructure investment2
. The majority of this infrastructure will be housed in the United States, positioning Anthropic to meet what CFO Krishna Rao described as "exponential growth" in the company's customer base while enabling Anthropic Claude AI models to "define the frontier of AI development"1
.
Source: DT
The deal gives Anthropic access to Google Cloud TPUs, or Tensor Processing Units, Google's advanced AI chips that compete directly with Nvidia's dominant GPUs in the semiconductor industry
5
. According to Broadcom's securities filing, the multi-year supply agreement runs through 2031, with Broadcom providing networking and other components for Google's next-generation AI racks2
. The filing notably states that Anthropic's use of the expanded capacity is contingent on its continued commercial performance, suggesting Broadcom views the arrangement as carrying financial risk3
.Revenue Surge Drives Infrastructure Expansion Amid Soaring AI Demand
The infrastructure expansion comes as Anthropic's annualized revenue run rate has surpassed $30 billion, up dramatically from approximately $9 billion at the end of 2025
1
. This represents a more than threefold increase in just three months, driven largely by enterprise adoption of Claude AI models. The company now serves more than 1,000 business customers spending over $1 million annually, double the figure from February when Anthropic closed its $30 billion Series G funding round that valued the company at $380 billion2
.The hardware and infrastructure required to develop a single gigawatt of capacity costs an estimated $35 billion to $50 billion, with the bulk spent on chips
5
. This suggests Anthropic's total commitment to Google and Broadcom could run into hundreds of billions of dollars as the company races to secure computing resources for both AI model training and serving customer demand. In all, the deal would give Anthropic access to close to 5 gigawatts in new computing capacity over the coming years, according to sources familiar with the terms5
.
Source: The Next Web
Multi-Cloud Strategy and Potential Move to Design Its Own AI Chips
Despite the expanded Google partnership, Amazon Web Services remains Anthropic's primary cloud and training partner under Project Rainier, the Trainium 2-based supercluster in Indiana
2
. The new Google-Broadcom capacity sits alongside that arrangement rather than replacing it, as Anthropic continues to draw heavily on Nvidia GPUs through cloud services including AWS, Google Cloud, and Microsoft Azure2
. This multi-cloud approach allows Anthropic to "match workloads to the chips best suited for them"3
.In a development that could reshape its hardware strategy, Anthropic is exploring the possibility of designing its own AI chips, according to three sources cited by Reuters
4
. The plans remain in early stages and the company has yet to commit to a specific design or assemble a dedicated team, with Anthropic potentially deciding to only purchase AI chips rather than design them4
. These discussions mirror similar efforts at Meta and OpenAI, though designing an advanced AI chip can cost roughly half a billion dollars as companies need skilled engineers and must ensure the manufacturing process has no defects4
.Related Stories
Industry Implications and Competitive Landscape
The Anthropic deal represents a significant win for Broadcom, which acts as the silicon implementation partner for Google, converting the search giant's TPU architecture into manufacturable ASIC layouts while supplying high-speed SerDes, power management, and packaging
2
. TSMC handles fabrication, while Google owns both the TPU architecture and software stack. Broadcom shares rose almost 3 percent after markets closed following the announcement5
. The same division of labor underpins Broadcom's separate $10 billion custom silicon program with OpenAI, announced as a 10 gigawatt co-development effort last October, making Broadcom the implementation layer for two of the three largest U.S. frontier model developers2
.
Source: Tom's Hardware
Analysts at Mizuho estimated that Broadcom would record $21 billion in AI revenue from Anthropic in 2026 and $42 billion in 2027, though the SEC filing didn't contain specific amounts
2
. These massive compute deals have drawn scrutiny for their circularity, with Big Tech groups like Google acting simultaneously as customers, suppliers, and investors in AI labs5
. Google has invested billions into Anthropic, giving it a 14 percent stake as of March last year. The arrangements reflect an industry-wide scramble to secure AI chip capacity amid an ongoing AI chip shortage, with OpenAI having struck similar computing deals with Broadcom, Nvidia, AMD and others to lock in as much capacity as possible5
.References
Summarized by
Navi
[1]
[2]
[3]
Related Stories
Anthropic Builds In-House Silicon Team for Custom AI Chips to Power Claude Models
05 Aug 2026•Technology
Anthropic enters talks with Samsung to develop custom AI chip, challenging Nvidia's dominance
02 Jul 2026•Technology
Anthropic explores Microsoft's Maia 200 AI chip to reduce Nvidia dependence amid compute struggles
21 May 2026•Technology
Recent Highlights
1
OpenAI Claims AI Math Breakthrough on Navier-Stokes, Sparks Controversy Over Credit and Future of Research
Science and Research
2
Anthropic Researcher Quits, Warns Self-Improving AI Poses Existential Risk to Humanity
Policy and Regulation
3
Nvidia CEO Jensen Huang Declares AGI Has Arrived With OpenAI's GPT-6 Astra Release
Technology
Recent Highlights
Today's Top Stories
Don’t drown in AI news. We cut through the noise - filtering, ranking and summarizing the most important AI news, breakthroughs and research daily. Follow topics that matter to you and stay ahead.
