Synopsys CEO Stakes Claim to Indispensability Across AI, Custom Silicon, and Foundry Workflows
Mizuho Technology Conference, June 9, 2026
Synopsys Chief Executive Sassine Ghazi delivered a forceful message at the Mizuho Technology Conference that borders on the definitive: the company has become structurally embedded in the semiconductor industry's most critical growth vectors. Speaking candidly about everything from the Ansys integration to concerns about Intel and the emerging agent-driven design paradigm, Ghazi repeatedly emphasized that "you cannot design these advanced chips without Synopsys" and went further to state that custom silicon for hyperscalers "cannot happen without Synopsys IP."
The company reported a strong fiscal second quarter that beat guidance on revenue, earnings per share, and operating margin, prompting management to raise full-year targets across all metrics including free cash flow. This marks the third quarter since closing the transformative Ansys acquisition, which Ghazi characterized as dramatically expanding the company's addressable opportunity.
Dual Factory Model Promises Material IP Revenue Upside
The most significant strategic development centers on Synopsys fundamentally restructuring its interface IP business around two distinct operating models. The traditional "factory one" approach of building standard IP once and selling it many times will continue serving the merchant chip market with expected mid-teens growth. But the company is now constructing what Ghazi calls "factory two" specifically for custom silicon opportunities at hyperscalers and other advanced customers requiring IP customization.
This distinction matters because hyperscalers building their own chips need customized IP delivered before standards are finalized, a capability Ghazi insists only Synopsys possesses at scale. The company is actively negotiating to shift from pure licensing fees to a hybrid model incorporating royalties on these custom engagements. Ghazi emphasized the royalty upside "should be fairly meaningful in order to justify even us talking about it or building that factory," though specific financial targets will be disclosed at the September 30 Investor Day.
The IP business bottomed in the first fiscal quarter and delivered sequential growth in the second quarter, with management committed to continued sequential improvement through year-end. Beyond the dual factory model, fundamental demand remains strong as chip starts proliferate across the industry, particularly for AI infrastructure applications where design activity continues accelerating.
Core EDA Growth Obscured by Accounting Complexities
Core EDA growth has appeared to soften below double digits in recent quarters, but Ghazi argues the proper measurement is trailing twelve months rather than sequential comparisons due to the lumpy nature of hardware revenue recognition. On that basis, the first half of fiscal 2026 delivered double-digit growth. The Ansys integration has created additional comparison difficulties because Synopsys divested certain technologies and keeps Ansys EDA separate from legacy Synopsys EDA to avoid double-counting when benchmarking against market data.
Management remains committed to long-term double-digit EDA growth driven by increasing design complexity and the industry's structural shift from monolithic chips to 3D integrated circuits. The latter trend directly supports the Ansys acquisition thesis, as stacking chips vertically introduces mechanical and thermal challenges that require multiphysics simulation alongside traditional electronic design automation.
Agent-Driven Design Creates New Monetization Layer
Synopsys released products in March featuring multi-agent adaptive orchestration with a cognitive layer calling specialized agents to move from specification to verified RTL in the front-end design process. As customers adopt hybrid workforces combining human engineers with what Ghazi terms "AgentEngineers," the company sees a new consumption-based revenue opportunity layered on top of traditional per-seat subscription licensing.
The business model evolution extends to discussions with AI labs interested in fine-tuning foundation models for semiconductor applications. In those scenarios, Synopsys is negotiating percentage-of-token revenue sharing arrangements, confident that its physics-based solvers will run underneath regardless of which model customers employ. Ghazi framed this as addressing the industry's persistent talent bottleneck, enabling customers to design more chips with existing headcount while reducing time-to-market and development costs.
Intel Relationship Remains Structurally Secure
Addressing persistent investor concerns about potential share loss at Intel following Lip-Bu Tan's appointment as CEO and the hiring of former Cadence executives, Ghazi was unequivocal: "You cannot design those chips without Synopsys." He characterized the relationship with Intel's new leadership as very strong with zero indication of any strategic shift away from Synopsys as primary partner, a designation dating back to 2005 or 2006 when Ghazi himself sponsored the relationship early in his career.
Looking at actual market share data rather than speculation, Ghazi sees no change whatsoever. On the foundry side, his language became even more emphatic, stating twice that "you cannot be in the foundry business without Synopsys" because the company serves as the essential bridge for onboarding customers to any foundry through its IP and foundation technologies. The multi-year engagement on Intel 18A continues with ongoing IP development work.
Ansys Integration Delivering Early Cross-Sell Wins
The first wave of joint Synopsys-Ansys products launched at the company's Converge event in March, fusing Ansys multiphysics simulation into Synopsys digital design flows for 3D-IC applications. Multiple customers are evaluating the technology with several expected to move into production deployment. Because these joint solutions represent new SKUs priced independently from standalone Synopsys or Ansys products, even existing customers of both companies will generate incremental revenue, delivering the "one plus one greater than two" outcome investors seek.
Ghazi drew parallels to the earlier Fusion Compiler product that integrated timing and power sign-off into the design platform, which not only commanded pricing uplift but drove market share gains because customers perceived superior value compared to discrete point tools. He expects the same dynamic with Ansys integration, particularly as customers designing AI chips confront thermal and structural challenges from advanced packaging and chiplet architectures.
The company's three-layer partnership with NVIDIA spans GPU-accelerated compute at the base, agentic orchestration in the middle, and physical AI at the top through Ansys integration with Omniverse. Ghazi clarified that NVIDIA's physics simulation capabilities target real-time gaming and visualization, while Ansys owns the sign-off physics simulation market requiring highest fidelity for actual manufacturing decisions, making the relationship complementary rather than competitive.
Margin Expansion Path Ahead of Schedule
Operating margin will expand more than 300 basis points this fiscal year through portfolio pruning, the previously announced 10% workforce reduction, and accelerating the $400 million Ansys cost synergy target from year three into earlier periods. Management remains committed to reaching mid-40s operating margin, acknowledging the company has trailed its closest peer but noting the 20-plus month regulatory approval process for Ansys limited flexibility to optimize the business during that extended period.
Design start data Synopsys tracks across the industry shows continued increases at companies serving AI infrastructure, with some customers expanding from four annual chip starts to five or six. The rest of the semiconductor market saw declining design activity that appears to have bottomed in the last two quarters, though the recovery has not yet materialized. Net aggregate chip starts are growing but not at the accelerated rate of pure AI infrastructure plays.