Synopsys launches agentic AI platform for chip and systems design

Synopsys, Inc. (Nasdaq: SNPS) announced a portfolio of domain-specific AI agents and a supporting platform aimed at automating engineering workflows across semiconductor and systems design, according to a company press release dated Sept. 28, 2026.

The company introduced AgentEngineer solutions, a set of long-horizon AI agents designed to handle tasks across verification, implementation, analog, manufacturing, and simulation domains. These agents operate on the Synopsys Autopilot Platform, which provides orchestration, memory, security controls, and workflow governance.

The platform is described as open, supporting integration with third-party models, infrastructure, and tools alongside Synopsys's own offerings. Security features include access controls, encryption, and runtime guardrails.

Synopsys cited results from early customer engagements, including up to 50x faster verification closure, 20% higher coverage, and a 30% productivity boost. Fujitsu's Senior Vice President Toshio Yoshida said his team achieved "a 10–30% productivity boost in RTL code generation" using the Verification AgentEngineer. Intel Vice President Manoj Selva said the technology showed "promise of agentic AI for debug, enabling engineers to accelerate issue resolution."

Partners mentioned in the announcement include Nvidia, Samsung Electronics, TSMC, MediaTek, Fujitsu, Intel, and AheadComputing. Nvidia's involvement includes accelerated computing infrastructure and its Agent Toolkit, Nemotron models, and OpenShell secure runtime.

Synopsys said more than 50 customer engagements are currently underway. General availability of the AgentEngineer solutions and Autopilot Platform is planned for the end of 2026.

You May Also Be Interested In

• Scorpio Gold Corp joins Terminal1 investor relations platform
• Anthropic in talks with neocloud Nscale over second multibillion-dollar deal - CNBC
• Palo Alto Networks (PANW) Reiterated at Buy by Stifel: 'We Favor Select Platform Vendors'