🔍 Read the full analysis: OpenAI May Buy Google TPUs: What It Could Mean For Nvidia’s Hold On AI Chips on ThorstenMeyerAI.com
Get movie nights delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
The Information and Bloomberg reported that OpenAI plans to buy Google’s tensor processing units, but the companies have not confirmed a deal and its terms are unknown. OpenAI is already pursuing several non-Nvidia chip arrangements, while Google has begun supplying TPUs beyond its own data centers. Any effect on chip prices or ChatGPT availability remains unclear.
The Information and Bloomberg reported this week that OpenAI plans to buy Google’s tensor processing units, or TPUs, potentially adding a new supplier as the ChatGPT maker expands its computing capacity. Neither company has confirmed a deal, and no price, chip quantity or delivery schedule has been disclosed.
The Information reported the plan on Oct. 27, citing one person familiar with the matter. Bloomberg reported it a day later, also citing one person. Reuters, relaying The Information’s report, said the financial terms and other details were not available. OpenAI declined to comment, while Google did not immediately respond to Reuters. Bloomberg’s reported account said final terms were not known.
There is evidence that OpenAI is working on TPU technology, but it does not establish that the company has placed an order. An OpenAI careers listing published on Oct. 20 sought a software engineer for TPU compilers, asking the hire to lead optimization of the company’s frontier models on TPUs. The posting indicates technical work; it is not a purchase announcement.
Google has begun making TPUs available beyond its own data centers. At its Oct. 28 earnings call, the company said nearly 100 customers use TPUs and named Anthropic and Apple among them. Oracle has separately announced plans to offer Google TPUs through its cloud from 2027. These confirmed moves provide context for the OpenAI report, but do not verify it.
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 Chip Options for OpenAI
If a purchase is completed, it would give OpenAI another route to secure computing capacity for training and running AI models. Demand for advanced chips has outstripped supply, and the company has been arranging access to hardware from several providers. A broader supplier base could give OpenAI more flexibility in where it runs workloads and how it negotiates pricing and delivery.
That would not by itself end Nvidia’s strong position. Nvidia remains a central supplier of AI accelerators, and a reported TPU plan says nothing yet about how many chips OpenAI might use or which workloads they would handle. The more immediate industry signal is that Google is trying to sell its own AI hardware to customers outside its cloud, including companies that compete with it.
For customers and ChatGPT users, any practical effect is uncertain. More available computing capacity could support OpenAI’s future services, but there is no confirmed link between a possible TPU purchase and lower prices, faster responses or a change to ChatGPT. The reported arrangements and other chip plans largely concern capacity expected in 2027 or later.
Popular right now
TCL T7 Series 85-Inch 4K 144Hz QLED Smart Google TV
Breville Bambino Espresso Machine, Brushed Stainless Steel (BES450BSS)
Vitamix 5200 Professional-Grade Blender, 64 oz. Container, Black
As an affiliate, we earn on qualifying purchases.
OpenAI’s Expanding Chip Portfolio
The reported Google plan would follow a series of public agreements that show OpenAI looking beyond Nvidia. The source material lists announced arrangements involving AMD, Broadcom, Cerebras and Amazon, alongside a 2025 letter of intent involving Nvidia. The figures described include 6 gigawatts for AMD, 10 gigawatts of custom chips with Broadcom, 750 megawatts with Cerebras and about 2 gigawatts of Amazon Trainium. These announced plans have different partners and terms; they should not be read as evidence that all the capacity is already delivered or in use.
OpenAI’s relationship with Google has also included earlier, limited TPU work. Reuters reported in June 2025 that OpenAI had started renting Google TPUs through Google Cloud. OpenAI then told Reuters it was in early testing with some TPUs and had no active plans to deploy them at scale. On Oct. 21, 2026, OpenAI said its frontier models would be available through Google Cloud, another indication of commercial ties between companies that also compete in AI.
Google’s shift to supplying customers directly marks a change in how it sells the hardware. At an April 2026 earnings call, Chief Executive Sundar Pichai said the company would begin delivering TPUs in hardware configurations to select customers’ own data centers, citing demand from AI labs and other users. Google has also announced cloud partnerships, including Oracle’s planned TPU offering from 2027.
“OpenAI, a Longtime Nvidia Customer, Plans to Buy Google’s AI Chips”
— The Information, in its Oct. 27 report
Deal Terms Remain Unreported
No purchase has been confirmed by OpenAI or Google. The reports rely on one source each, and the companies have not publicly detailed whether discussions have led to a binding agreement. The available reporting does not establish the number or generation of TPUs, the purchase price, delivery dates, or where the hardware would operate.
It is also unclear whether any chips would be installed in OpenAI facilities, accessed through Google’s cloud, or supplied through a partner such as Oracle. The TPU compiler job posting and earlier testing support the view that OpenAI has technical interest in the hardware, but neither proves a purchase or a large-scale deployment.
OpenAI’s other chip agreements also have varying terms and timelines. The announced capacity figures do not establish how much hardware is available now or how it will be allocated. No information in the source material shows whether a possible Google arrangement would replace, supplement or simply test alongside other suppliers.
Watch for Terms and Deliveries
The next confirmation would need to come from OpenAI, Google or a formal filing or agreement that identifies the transaction. Until then, the reported plan should be treated as unconfirmed, rather than a completed purchase. Further reporting may clarify the price, quantity, chip model, deployment location and timing.
Google’s planned Oracle cloud offering is scheduled to begin in 2027, while the source material says many of OpenAI’s other chip arrangements point to capacity in 2027 or later. Those milestones may offer a clearer view of how quickly alternative hardware becomes available. For now, there is no announced change to ChatGPT and no confirmed user-facing impact.
Key Questions
Has OpenAI confirmed that it will buy Google TPUs?
No. The Information and Bloomberg reported that OpenAI plans to buy the chips, but neither OpenAI nor Google has confirmed a deal.
What are Google TPUs?
TPUs, or tensor processing units, are AI chips designed by Google. The company uses them in its data centers and has begun supplying them to selected outside customers.
Would a TPU deal end OpenAI’s reliance on Nvidia?
There is no evidence that it would. OpenAI has announced or reported arrangements involving several chip suppliers, but the reported Google plan’s size and role are unknown. Nvidia remains a major supplier of AI chips.
Could the report change ChatGPT for users?
No user-facing change has been announced. A potential purchase could add computing capacity, but any effect on ChatGPT’s speed, availability or price is unknown.
When could Google TPUs become available through Oracle?
Oracle has announced plans to offer Google TPUs through its cloud from 2027. That timeline does not confirm when or whether OpenAI would use them.
Source: ThorstenMeyerAI.com
