🔍 Read the full analysis: OpenAI May Buy Google’s TPUs: What Could Change In The AI Chip Market? on ThorstenMeyerAI.com
Get business pricing on tech for your team
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
The Information and Bloomberg reported that OpenAI plans to buy Google’s AI chips, citing one person familiar with the matter. Neither company has confirmed a deal, and its size, price and timing are unknown. If it proceeds, the move would add to OpenAI’s growing list of non-Nvidia chip arrangements and give Google another route to sell TPUs.
The Information and Bloomberg reported on October 27 and 28 that OpenAI plans to buy Google’s tensor processing units (TPUs), chips used to run artificial intelligence workloads. Neither company has confirmed a deal, and no price, chip quantity or delivery schedule has been disclosed. If the reported plan goes ahead, it would add Google to OpenAI’s expanding group of chip suppliers and mark a potential sale by a major AI rival to one of its competitors.
Both reports cited one person familiar with the matter. Reuters, relaying The Information’s report, said the financial terms and other details could not be learned. OpenAI declined to comment, while Google did not comment or did not immediately respond to inquiries, according to the accounts in the supplied source material. The reports do not establish that a purchase order has been signed.
There is evidence that OpenAI is working on TPU compatibility, but it does not confirm a purchase. Its careers page listed a TPU Compilers software engineering role, posted October 20, seeking someone to optimize its frontier models on TPUs. A job listing indicates technical work; it is not evidence of a purchase or large-scale deployment.
Google has begun selling TPU systems beyond its own data centers. On an April 2026 earnings call, CEO Sundar Pichai said the company would deliver TPUs in a hardware configuration to a select group of customers for use in their own data centers. On October 28, Google said nearly 100 customers use TPUs and named Anthropic and Apple, while announcing partnerships with Oracle and Crux AI. Oracle has said it plans to offer Google TPUs through its cloud from 2027.
Infographic 01 · Reported, documented, confirmed
OpenAI and Google’s TPUs: what we actually know
The purchase itself is a report. The pieces around it, Google selling its chips and OpenAI hiring for them, are on the record.
- OpenAI plans to buy Google’s AI chipsReported
The Information, 27 October; Bloomberg, 28 October. Each cites one person familiar with the matter. OpenAI declined to comment; Google did not comment.
- OpenAI is hiring to optimize its frontier models on TPUsPrimary document
“Software Engineer, TPU Compilers”, posted on OpenAI’s careers page on 20 October 2026. A sign of serious work, not a purchase order.
- Google sells TPU systems into customers’ own data centersConfirmed
Announced on 29 April, first deliveries in the second quarter, “a meaningful revenue ramp in 2027” (Alphabet’s CFO, 28 October).
- Oracle will offer Google TPUs in its cloudConfirmed
The first cloud outside Google to do so. Available in calendar year 2027, ramping over several years.
- Price, number of chips, TPU generation, delivery dates, locationNot known
Nothing published. Not even whether the chips would sit in Google’s data centers, OpenAI’s or a partner’s.
How OpenAI and Google got here
- Jun 2025OpenAI rents Google TPUs, calls it early testingReuters: no active plans to deploy them at scale
- 29 Apr 2026Google: TPUs for customers’ own data centersAlphabet Q1 earnings call
- 20 Oct 2026OpenAI posts a TPU compiler jobOpenAI careers page
- 21 Oct 2026OpenAI models coming to Google CloudOpenAI announcement, as reported by Reuters
- 27 Oct 2026The Information reports the purchase plan; Oracle announces TPUsReported · Oracle press release
- 28 Oct 2026Bloomberg reports it too; Alphabet’s call does not mention OpenAIReported · Alphabet Q3 call
“Reported” means the outlets’ own reporting, which I could read only in part (paywalls). “Confirmed” means the company’s own statement or filing.
Sources: The Information via Reuters, 27 Oct 2026, Bloomberg, 28 Oct 2026, as cited by The Asia Business Daily, OpenAI careers, posted 20 Oct 2026, Reuters, 30 Jun 2025, Alphabet Q1 call, 29 Apr 2026, Q3 call, 28 Oct 2026, Oracle, 27 Oct 2026.
Powered by Thorsten Meyer AI
More Options for OpenAI’s Compute
If the reported deal is completed, it would give OpenAI another option for the chips that power AI training and services. The potential impact is not that Nvidia is suddenly displaced: Nvidia remains a major supplier, and the report provides no information about the scale of any Google purchase. Rather, a credible ability to allocate some workloads to different hardware could give OpenAI more flexibility as it builds computing capacity.
That flexibility could matter in negotiations over price, delivery and access to computing resources. This is a possible market effect, not a confirmed outcome of the reported plan. There is no disclosed agreement or evidence that a TPU purchase would reduce OpenAI’s costs, change ChatGPT performance or pass savings on to users. For customers and users, the immediate takeaway is limited: the development is a report about a possible supply arrangement, not a change to an AI product.
The deal would also carry a competitive complication. Google competes with OpenAI in AI products while also selling infrastructure that OpenAI could use. The relationship is not purely hypothetical: OpenAI said on October 21 that it would make its frontier models available through Google Cloud. Buying or renting a rival’s hardware can broaden supply choices, but the practical effects depend on terms and how much computing work moves onto it.
Products Worth Considering
Apple AirPods Pro 3 Wireless Earbuds with Active Noise Cancellation
LG UltraGear 34GX90SA-W 34-inch WQHD Curved OLED Gaming Monitor
Jackery Explorer 2000 v2 Portable Power Station with Two 200W Solar Panels
As an affiliate, we earn on qualifying purchases.
Google Opens TPU Sales Beyond Cloud
Google designed TPUs for its own data centers and has also made them available to cloud customers through Google Cloud. The 2026 move to deliver systems to selected customers’ own data centers expands that business. Google’s announcement of Oracle availability from 2027 adds another route to the chips, though it does not establish when or whether OpenAI would receive them.
OpenAI has already announced several non-Nvidia chip arrangements, so a Google deal would not be its first move beyond Nvidia. The source material lists announced plans involving AMD, Broadcom, Cerebras and Amazon, alongside a 2025 letter of intent involving Nvidia. It describes figures of 6 gigawatts for AMD, 10 gigawatts for Broadcom custom chips, 750 megawatts for Cerebras and about 2 gigawatts for Amazon Trainium, as well as 10 gigawatts in the Nvidia letter of intent. These are announced commitments or plans, not proof that all the capacity has been delivered or is operating.
OpenAI also had earlier, limited exposure to TPUs. Reuters reported in June 2025 that the company had started renting them through Google Cloud. OpenAI told Reuters at the time it was in early testing and had no active plans for large-scale deployment. That history supports the distinction between testing a chip and committing to buy it at scale.
““As TPU demand grows from AI labs, capital markets firms, and high performance computing applications, we will begin to deliver TPUs to a select group of customers in their own data centers in a hardware configuration.””
— Google CEO Sundar Pichai, on the company’s first-quarter 2026 earnings call
Deal Size and Delivery Remain Unknown
The reports remain unconfirmed by both companies. The source accounts cite one person familiar with the matter apiece, and the available information does not establish whether the two outlets relied on the same source. No public document or company statement in the supplied material confirms a signed deal.
Key details are missing: the number and generation of TPUs, price, delivery dates, contractual terms and where the chips would run. They could operate in Google’s data centers, in OpenAI’s facilities or through a cloud partner; the reporting described here does not settle that question. It is also unclear whether OpenAI would buy hardware outright, obtain capacity another way, or proceed beyond testing and technical work. Any effect on Nvidia’s market position, OpenAI’s costs, model performance or consumer prices is likewise unestablished.
Watch for Terms and Deployment
The next meaningful evidence would be a confirmation or further reporting that establishes the agreement’s terms, including scale and timing. OpenAI and Google could clarify whether the plan is a direct hardware sale, a cloud arrangement or another form of access. Until then, the reports should be treated as a reported intention rather than a completed transaction.
Oracle’s stated plan to offer Google TPUs from 2027 is a separate, announced development that may show how broadly Google can distribute the chips. For OpenAI, any later disclosure about deployment would help distinguish experiments and engineering preparation from sustained use. No change to ChatGPT or other OpenAI services is established by the report, and the available information gives users no confirmed product change to expect now.
Key Questions
Has OpenAI confirmed it will buy Google TPUs?
No. The Information and Bloomberg reported the plan, each citing one person familiar with the matter. The source material says OpenAI declined to comment and Google did not comment. No confirmed purchase is established.
What are Google TPUs?
TPUs, or tensor processing units, are chips Google designs for AI and other computing workloads. Google has used them in its data centers and is expanding sales to outside customers.
Would this be OpenAI’s first non-Nvidia chip arrangement?
No. The source material lists announced arrangements involving AMD, Broadcom, Cerebras and Amazon, in addition to OpenAI’s relationship with Nvidia. A Google arrangement would add another supplier, not mark OpenAI’s first step beyond Nvidia.
Could Google TPUs make ChatGPT cheaper or faster?
There is no confirmed evidence that the reported plan would change ChatGPT’s price or speed. Those outcomes would depend on whether a deal is completed, how many chips are involved and how OpenAI uses them.
When could OpenAI receive the chips?
No delivery date has been reported for OpenAI. Oracle separately plans to offer Google TPUs through its cloud starting in 2027, but that schedule does not establish when or how OpenAI might access them.
Source: ThorstenMeyerAI.com
