--- title: "Nvidia and Mistral Will Publish the AI Benchmarks That Can Reprice Dassault and Cadence's Software" description: "Mistral and Nvidia plan open benchmarks for engineering simulation. If toolmakers publish the time and compute a result takes, the per-seat software license loses its cover." publisher: "The Inference" section: "Platforms" published: 2026-10-22T19:00:50.230Z modified: 2026-10-22T19:00:50.230Z canonical: https://theinference.org/article/nvidia-and-mistral-will-publish-the-ai-benchmarks-that-can-reprice-dassault-and-cadence-s-software language: en keywords: "Simulation Software, Mistral AI, Nvidia, EDA, Artificial Intelligence" --- # Nvidia and Mistral Will Publish the AI Benchmarks That Can Reprice Dassault and Cadence's Software > Mistral and Nvidia plan open benchmarks for engineering simulation. If toolmakers publish the time and compute a result takes, the per-seat software license loses its cover. [Mistral AI](https://theinference.org/markets/companies/mistral-ai), the French open-weight lab, and [NVIDIA](https://theinference.org/markets/companies/nvidia) said on October 21 that they will run engineering simulation as a closed loop, pairing [Mistral Large 4](https://theinference.org/article/mistral-prices-large-4-at-a-twelfth-of-openai-s-gpt-6-astra-rate-as-a-langchain-test-cut-coding-costs-64) with NVIDIA's CUDA-X, Warp and Omniverse libraries so agents design, simulate and iterate without a person routing every step [1]. The promise is speed; nothing measured backs it yet. The commercial event is the other half of the announcement: public benchmarks and open datasets that score AI agents on accuracy, completed workflow, execution time and compute cost across computer-aided engineering and the chip-design software known as EDA [1]. Whoever writes the yardstick industrial buyers use to judge engineering AI takes a share of the pricing power that simulation vendors have collected per seat for decades. ## The physics work is real; the measured results are not Mistral has spent the year building the middle hops between a model and an engineer's desk. It bought the physics-modeling startup Emmi AI in May, opened a Munich hub in September with BMW and [Siemens Energy](https://theinference.org/markets/companies/siemens-energy) as industrial partners, and signed a statement of intent with CERN, the European particle-physics laboratory, on October 15 to train a foundation model on simulated collider data [2][3]. NVIDIA's partnership is the commercial front, opening a week after [Germany's SAP-Mistral pact proved to bind no one](https://theinference.org/article/germany-says-its-sap-mistral-ai-pact-bound-no-one-sap-shares-are-up-57-into-october-21-earnings) [2][3]. The target is a genuine bottleneck, because a single simulation run can take days [2]. What does not exist yet is evidence on the new integrations: the companies have announced no release dates and no measured speed or accuracy gains [4]. Everything past the announcement, the benchmarks included, is a plan. ## The incumbents already run on NVIDIA's stack NVIDIA's same-day update from its Berlin conference shows this is not a two-camp fight. Siemens estimates 20% cost savings from agents built on NVIDIA's Factory Operations Blueprint [5]. imec, the Belgian microelectronics research center, reports its lithography simulations running up to 40 times faster [5]. [Dassault Systèmes](https://theinference.org/markets/companies/dassault-systemes), the French simulation-software vendor, runs design companions powered by Nemotron, NVIDIA's own open model family, on CUDA-X-accelerated solvers, and [Cadence](https://theinference.org/markets/companies/cadence), the [chip-design software vendor](https://theinference.org/article/amazon-s-aws-pays-synopsys-over-1-billion-for-chip-blueprints-to-build-past-nvidia), is benchmarking its fluid-dynamics workloads on NVIDIA GPUs [5]. [Synopsys](https://theinference.org/markets/companies/synopsys), which has absorbed the simulation leader Ansys, signed its own [frontier-model deal with OpenAI](https://theinference.org/article/altman-calls-cerebras-a-close-partner-as-openai-s-fastest-tier-runs-on-nvidia) on September 30, with revenue sharing attached [6]. NVIDIA supplies both camps, and it holds a Mistral stake alongside [ASML](https://theinference.org/markets/companies/asml), the Dutch chip-equipment maker [7]. The co-author of the yardstick earns more each quarter than the listed simulation vendors earn in a year [9][10][11]. *Chart: **The yardstick's co-author dwarfs the listed simulation vendors** Bar chart comparing latest completed fiscal-year revenue: Nvidia about $216 billion, Synopsys about $7.1 billion, Cadence about $5.3 billion* *Annual revenue, most recent completed fiscal year; fiscal year-ends differ (Nvidia January 2026, Synopsys October 2025, Cadence December 2025). SEC filings, USD billions. [9][10][11]* ## The first benchmark release is the date to watch For buyers, a public yardstick means comparing approaches without taking vendor demonstrations on faith, and the research findings behind the announcement name Airbus, BMW and ASML as clients that stand to benefit [8]. For the toolmakers, the benchmarks will publish the time and compute a correct result takes, which is exactly what a per-seat license has priced without saying so [1]. That pressure on pricing is a projection; every benchmark release that publishes those numbers pushes the same way. When the compute cost of a validated simulation becomes a public number, the per-seat license loses its cover. The reward belongs to whoever does the unglamorous middle work between capability and desk, and this month that is Mistral's physics team in Munich. ## Takeaway By creating open benchmarks for physics and chip simulation, Nvidia and Mistral are turning engineering runtimes into public metrics, threatening legacy per-seat pricing across CAE and EDA software. ## Sources 1. [Mistral AI, "Mistral Advances AI for Engineering with NVIDIA", 21 October 2026](https://mistral.ai/news/mistral-nvidia-ai-for-industrial-engineering/) 2. [Unite.AI, "Mistral AI Opens Munich Hub Focused on Physics and Industrial AI", 28 September 2026](https://www.unite.ai/mistral-ai-opens-munich-hub-focused-on-physics-and-industrial-ai/) 3. [The Next Web, "CERN and Mistral plan an AI foundation model for particle physics", 16 October 2026](https://thenextweb.com/news/cern-mistral-large-physics-model) 4. [TokenPost, "Mistral AI, NVIDIA Expand Partnership Into Engineering Simulation", 21 October 2026](https://www.tokenpost.com/news/technology/37088) 5. [Blockchain News, "NVIDIA Details Industrial AI Push Across Europe's Manufacturing Base", 21 October 2026](https://blockchain.news/news/nvidia-europe-industrial-ai-manufacturing-cuda-x-omniverse-cosmos-nemotron) 6. [Synopsys news release, "OpenAI and Synopsys Announce GPT-Synopsys: Frontier Intelligence to Revolutionize Chip Design", 30 September 2026](https://news.synopsys.com/2026-09-30-OpenAI-and-Synopsys-Announce-GPT-Synopsys-Frontier-Intelligence-to-Revolutionize-Chip-Design) 7. [Forbes, Ron Schmelzer, "AI Is Moving From Writing Code To Designing Physical Products", 21 October 2026](https://www.forbes.com/sites/ronschmelzer/2026/10/21/ai-is-moving-from-writing-code-to-designing-physical-products/) 8. [Crypto Briefing, "Mistral and Nvidia team up to bring AI into engineering simulation", 21 October 2026](https://cryptobriefing.com/mistral-nvidia-ai-engineering-simulation/) 9. [Nvidia, annual figures from its SEC filings, retrieved 22 October 2026](https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001045810) 10. [Synopsys, annual figures from its SEC filings, retrieved 22 October 2026](https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0000883241) 11. [Cadence, annual figures from its SEC filings, retrieved 22 October 2026](https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0000813672) --- Nvidia and Mistral Will Publish the AI Benchmarks That Can Reprice Dassault and Cadence's Software — The Inference. Canonical: https://theinference.org/article/nvidia-and-mistral-will-publish-the-ai-benchmarks-that-can-reprice-dassault-and-cadence-s-software