Synopsys (Nasdaq: SNPS) announced two structurally different AI deals on 30 September 2026 — one with Amazon, one with OpenAI. They are not the same deal, and the difference is the story.
GPT-Synopsys is an announced intention, not a shipping product. Every capability below is described in future tense, and the release gives no availability date and no price.The revenue sharing arrangement is confirmed to exist and described no further: no split, no percentage, no dollar value. The separate Synopsys agreement with Amazon announced the same day is a different instrument and its billion-dollar figure does not belong to this one. Nothing here is investment advice.
Synopsys announced two separate multi-year agreements on the same day. The first, with Amazon, is valued at more than one billion dollars and includes a royalty tied to Amazon’s chip output. The second, with OpenAI, carries a revenue-sharing arrangement and a go-to-market collaboration. No dollar figure, split, or percentage is attached to the OpenAI agreement anywhere in the release.
The OpenAI agreement centers on a jointly developed model called GPT-Synopsys. It is purpose-built to operate Synopsys electronic design automation tools. The release names Sassine Ghazi, Synopsys’ president and chief executive, and Greg Brockman, OpenAI’s president and co-founder, as the principals behind it.
The bundle is specific. The release states: “The joint service offering will provide the bundled compute, model, and licenses.” Compute and model come from OpenAI. GPT-Synopsys runs on OpenAI-hosted infrastructure and integrates with Synopsys.ai and the Synopsys Autopilot platform. Synopsys supplies what Ghazi calls “the underlying, ground-truth engineering tools.”
The key insight: These are two different commercial instruments. The Amazon deal is a royalty on a customer’s silicon. The OpenAI deal is a revenue share on a jointly built product. Merging them — or attaching the billion-dollar figure to OpenAI — misreads what Synopsys actually signed.
The components are the release’s own words. What the chart records is which party supplies each, and that the commercial terms binding them are not published.
The Structural Read
An EDA company’s traditional moat is the tool plus the licence that gates access to it. This agreement puts both inside someone else’s product. That is a real shift in position — and it may also be a smart trade.
Ghazi’s own language places Synopsys deliberately. He describes the tools as “underlying” and “ground-truth.” Underlying is a position in a stack — one layer below the surface the engineer actually touches. In this bundle, OpenAI owns that surface.
But the logic runs both ways. A model that is purpose-built to operate Synopsys tools — and that only works properly against them — makes those tools harder to replace. A revenue share pays Synopsys on usage it does not have to sell, provision, or support itself. The company that loses the interface may gain a distribution channel it could not have built alone.
Greg Brockman, OpenAI President & Co-Founder
“Better chips make better AI.”
Brockman’s framing is candid about OpenAI’s motive. This is not a services play. OpenAI needs better chips to run larger models. Embedding itself into the design workflow is a way to accelerate the hardware that feeds its own ambitions. Synopsys gets distribution and recurring revenue. OpenAI gets influence over the silicon layer.
The distinction that matters most: the company that owns the interface owns the renewal conversation. In this bundle, that is OpenAI. Whether that is a concession or simply a new channel depends entirely on how the revenue share is structured — terms that are not public.
What the Release Does Not Say
Read the tenses carefully. The release says GPT-Synopsys “will run” on OpenAI-hosted infrastructure, “will be designed to interoperate” with customer harnesses, “will be deeply integrated” with Synopsys.ai, and that the offering “will provide” the bundle. Every capability sits in that tense. The product described is an intention, not something a customer can purchase today.
The release’s own forward-looking-statements section confirms this. It lists the development, capabilities, timing, availability, and anticipated benefits of GPT-Synopsys among the things that could differ materially from what is described. That is standard practice for a partnership announcement. It is worth stating plainly because the headline reads like a launch.
What is absent: the revenue-share split, any dollar value for the OpenAI agreement, the term length, pricing, a general-availability date, and the identity of any early customer. Early engagements with leading semiconductor customers are described as underway. None are named, and no count is offered.
What It Changes
Three readings follow, and the first is about how Synopsys gets paid.
A revenue share pays on usage the company does not have to originate or service, which is a different mechanism from selling seats. The Amazon royalty shows the same willingness to tie revenue to downstream activity. Whether either is expansion or dilution depends on terms that neither release gives.
The second is the change in how the tools get driven. The release draws the distinction itself: today agentic systems connect general-purpose models to EDA tools, and the stated next step is models that become expert users of them. Engineers delegate objectives such as power, performance and area targets, the PPA the industry optimises against, rather than individual commands, with agents handling timing and verification closure and surfacing results for review.
The third is about who owns the surface. OpenAI is not building EDA tools; it is building the interface through which engineers would use them. On the arrangement as described, the model layer sits between the engineer and the underlying tool, and the release says that layer runs on OpenAI-hosted infrastructure.
The data terms are the specific part. Customer data is not used to train the model, is encrypted at rest and in transit, and is subject to configurable retention, audit and permission controls.
Business Engineer Framework
FDE Framework: Founders, Distributors, Enablers
The GPT-Synopsys deal is a case study in what happens when an Enabler and a Founder co-develop a product. Synopsys supplies the ground truth; OpenAI supplies the surface. The FDE Framework maps exactly where each company sits in the AI stack — and why interface ownership is the variable that determines who controls the renewal conversation. The full Map of AI analysis shows how these positions are playing out across 200+ companies in 9 stack layers.
Synopsys closed 30 September 2026 with two structurally distinct deals: a billion-dollar-plus royalty arrangement with Amazon tied to chip output, and a revenue-sharing co-development agreement with OpenAI whose financial terms are not public. GPT-Synopsys is a real product with a real technical thesis — frontier models as expert users of EDA tools, delegating design objectives rather than commands — but it is not shipping, has no disclosed price, and has no named customers.
The bundle logic is sound: Synopsys supplies the ground truth, OpenAI supplies the interface, and a model that only works against Synopsys tools makes those tools harder to displace. Whether that trade is favorable depends entirely on the revenue-share terms, and those terms are not in the release.
91,000+ executives read Business Engineer for the AI strategy frameworks cited by ChatGPT, Claude, and Perplexity.
Every detail above comes from Synopsys’ own news release of 30 September 2026, read directly on news.synopsys.com. Nothing has been independently verified, and OpenAI published no separate statement that was read for this piece.The release confirms that the agreement includes a revenue sharing arrangement and says nothing further about it. No split, no percentage, no floor and no dollar value appears anywhere, and “multi-year” is the only duration given.
None has been supplied above.GPT-Synopsys is not a shipping product. Every capability in the release is described in future tense, there is no availability date, no price and no general-availability commitment, and the release’s own forward-looking-statements section lists the development, capabilities, timing, availability and anticipated benefits of GPT-Synopsys among the things that could differ materially from what is described.The separate Synopsys agreement with Amazon announced the same day is a different instrument: a multi-year arrangement worth more than a billion dollars that adds a royalty tied to the customer’s chip output.
The billion-dollar figure and the royalty belong to that agreement and not to this one.The observation that the joint offering places Synopsys’ licences alongside OpenAI’s compute and model, on OpenAI-hosted infrastructure, follows from the release’s own description. It is a structural reading and not a judgement about which party benefits. Not established and therefore absent: the revenue-share split, any dollar value, the term length, pricing, an availability date, and which customers are engaged or how many. Nothing above predicts anything, and nothing here is investment advice.
Gennaro is the creator of FourWeekMBA, which reached about four million business people, comprising C-level executives, investors, analysts, product managers, and aspiring digital entrepreneurs in 2022 alone | He is also Director of Sales for a high-tech scaleup in the AI Industry | In 2012, Gennaro earned an International MBA with emphasis on Corporate Finance and Business Strategy.
Scroll to Top
Discover more from FourWeekMBA
Subscribe now to keep reading and get access to the full archive.
