Cadence unveils Nvidia-backed AI models that chipmakers can run in-house

What happened

Cadence Design Systems launched Cadence Maestro, a suite of AI models built partly on Nvidia's open-weight Nemotron model and secured with Nvidia's OpenShell safety technology, to help design computer chips. Customers can run it on Cadence's cloud, public clouds, or entirely in-house.

Confirmed

Global impact / market context

Chipmakers can keep sensitive design data behind their own firewalls, which means they avoid sending valuable designs to outside cloud servers. Cadence says early results show up to 30% higher quality and up to five times better cost-efficiency than general-purpose models, which could make its tools more attractive.

Analyst inference

Cadence's in-house, model-agnostic approach contrasts with rival Synopsys, which said it would build GPT-Synopsys with OpenAI and run it on OpenAI-hosted infrastructure. The split matters because chip design software buyers may weigh data control against hosted convenience, shaping which vendor wins future contracts.

Analyst inference

What to watch

1. Watch whether Cadence shares customer adoption details for Maestro, since the company says customers can fine-tune models on their own chip designs and keep everything behind their own firewalls. Confirmed
2. Cadence could face competitive pressure if Synopsys's OpenAI-hosted GPT-Synopsys model proves easier or cheaper for chipmakers to deploy at scale. Proposed
3. Chipmakers may favor in-house tools if data control outweighs hosted convenience, which could shift design software revenue and spending toward Cadence and Nvidia. Analyst inference

Evidence

• Original source link unavailable