Synopsys is a formidable business that still deserves more investor attention. The company operates a mission-critical duopoly with Cadence in EDA (semiconductor design software), controlling roughly 65-70% of the EDA market between them. And mission-critical is truly mission-critical here – Nvidia, AMD, and Apple can’t design a chip without Synopsys or Cadence. Fabs like TSMC, Intel, and Samsung even require sign-off from Synopsys or Cadence tools before taking a chip into production.Furthermore, Synopsys revenue is overwhelmingly subscription- and royalty-based, with 75-80% recurring. Combine that with the mission-critical positioning, and you get an unusually reliable performer through semiconductor cycles. Synopsys has posted only three quarters of negative growth in fifteen years. And the moat protecting all of this is about as close to impenetrable as it gets: switching costs are prohibitive (entire verification methodologies and years of engineering muscle memory built around Synopsys’s specific tools), and decades of accumulated R&D in a highly complex industry combined with deep verification integration with fabs create extreme barriers to entry.That makes a business of exceptional quality, and one positioned for decades of compounding, especially as it is poised to benefit from AI accelerators, advanced packaging, chiplets, and shrinking nodes, all driving growing design complexity, adding to the usage intensity of EDA software and IP.Last week, the company released its fiscal Q3 results, which were impressive and sparked renewed optimism. The company beat consensus estimates and raised guidance, while providing very positive commentary on underlying trends and demand. Additionally, Synopsys is stepping up its rate of innovation – one of my primary points of criticism back in June – with a slate of big product releases and partnerships that can significantly contribute to growth in 2027. In response, shares gained 13% in Thursday’s trading session.Nonetheless, sentiment toward Synopsys remains somewhat poor, and shares are still down 28% over the last year, even after rebounding 19% from late July. You see, Synopsys has been slightly out of favor for most of the last year, mainly due to AI disruption concerns. This caught even more steam last quarter as news surfaced of a Chinese AI model reportedly capable of fully designing a chip with only freely available EDA tools, completely disintermediating the expensive Cadence and Synopsys software. It caused Synopsys to shed 9% of its value in a single day.Given these developments, I think it’s time to revisit this business and update my thesis. Therefore, in today’s post, I will go over the Q3 results and developments, break down the AI disintermediation risk, and ultimately update the medium-term outlook, my financial model, and my view of the stock.Without further ado, let’s delve in!This is a paid-exclusive analysis. Want full access? Consider upgrading to InvestInsights PRO for $9.50 a month or $95 a year! This gets you:6–8 Research Reports per month, 2x more than free subs do, including 3-4 Deep Dives and 2 mid-week analyses (that comes down to just $1 per analysis)Full visibility into my personal portfolio, including allocations, transactions, and conviction levelsAn overview of target prices and ratings, updated weeklyInvestInsights Premium is designed to help you find best-in-class compounders, identify deep-value opportunities, avoid costly mistakes, and stay focused on fundamentals when markets get noisy, with full transparency into how I make decisions over time.Subscribe nowFinancial & Performance ReviewSynopsys released its fiscal Q3 results last Thursday, August 26, and honestly delivered an excellent report that exceeded my expectations and comfortably cleared the Wall Street consensus. Synopsys showed improved momentum across the board and exceeded the high end of guidance on revenue, operating margin, and EPS, as headwinds finally eased and design momentum picked up.Starting at the top, Synopsys delivered Q3 revenue of $2.48 billion, clearing the consensus by $40 million and up 42% YoY, in line with Q2. However, this includes $711 million from the Ansys acquisition that closed in Q3 last year, as well as a headwind from the divesture of the Optical Solutions Group, so this isn’t apples-to-apples.Organic growth is likely closer to the 8-10% range, by my estimate, after adjusting for acquisitions and divestitures. That’s a solid performance, though still below what we have gotten used to from Synopsys and shy of management’s long-term targets, as it continues to work through near-term headwinds.Yet, more importantly, revenue came in ahead of expectations, and momentum improved significantly in Q3, with demand stronger than anticipated, especially in IP. The backlog (or RPO) also remains strong at $10.9 billion, modestly down sequentially due to the divestiture of the processor IP solutions business that closed last quarter. Breaking down revenue by segment, let’s start with Design Automation, which brought in roughly $2 billion in revenue, including the Ansys contribution, partially offset by the Optical Solutions Group divestiture. This includes 8.5% YoY EDA revenue growth, reflecting robust performance in EDA software and a record quarter for hardware-assisted verification solutions.This Q3 EDA growth is slightly above expectations and more impressive than it looks, given the tough comp, lapping 16% growth last year, which shows very strong design activity. The driver here is mainly AI-related, with design activity accelerating. Meanwhile, non-AI activity has been rough over the last year, but this now seems to be stabilizing.For perspective, Synopsys delivered somewhat depressed EDA growth for most of last year, largely because the U.S. government imposed a 6-week export ban on its operations in China in 2025. In itself, that 6-week ban wasn’t a huge drag, but the effects continued to linger due to changes in customer behavior. Simply, Chinese customers were hesitant to sign a large EDA contract with Synopsys, given their awareness of the risk of new restrictions that could significantly impact their chip development – losing access to Synopsys systems halfway through the chip development process would be detrimental, so this caution makes sense.This caused a major drag on Synopsys’ non-AI design activity, but it now seems to have bottomed out, allowing growth to recover and return to the norm. Additionally, back in June, I wasn’t as pleased with Synopsys’ EDA deal momentum compared to Cadence, but this has improved a lot in Q3.For starters, a year after completing the acquisition, Synopsys has now launched its first joint Synopsys and Ansys solution, Multiphysics Fusion. This combines Synopsys and Ansys technology and is the industry’s only solution with thermal analysis fully integrated into the chip design flow. Early adopters include Nvidia, Cisco, MediaTek, and Samsung Foundry, and these are seeing up to 10x faster design closure, 3x faster runtime, and as much as 86% IR fix rates.This is a great value-add. Management was clear that deployment is still in the early stages, but it expects strong adoption and the add-on to start contributing to growth in 2027, as adoption moves from validation into production. Above all, it is great to see Synopsys adding Ansys functionality to its product suite, which is a real edge over the competition. Given the value it brings, I expect it to contribute meaningfully to growth.On top of that, Synopsys showcased a fully autonomous, long-running design verification agent that can orchestrate the entire chip verification cycle and deliver up to 50x faster time to validated RTL while achieving a 20% coverage improvement during the DAC conference, partnering with Nvidia. These agentic AI solutions are already seeing strong customer interest, with more than 30 active customer engagements underway, including AMD and Microsoft.Finally, Synopsys also expanded its collaboration with Intel Foundry by certifying AI-powered EDA on Intel 14A nodes. This builds on Synopsys’ EDA flows and IP portfolio for Intel 18A and Intel 18A-P technologies.Above all, I am really pleased with the improved deal momentum and step-up in innovation cadence, which addresses some previous concerns. I expect these new products – Multiphysics Fusion and Agentic Design – to become meaningful contributors to growth as early as fiscal H1 2027, given the huge demand Synopsys is seeing and the significant design productivity step-up these products generate. This combination of new product launches, easing headwinds, and easier comps is already showing up in management’s guidance, which now points to further acceleration in EDA revenue growth in Q4 and double-digit FY26 revenue growth, suggesting a strong ramp from here.This is trending nicely ahead of my previous expectations.Moving to IP, Synopsys reported segment revenue of $474 million, up 11% YoY, which reflects a return to growth.You see, IP revenue growth turned deeply negative over the past year due to a number of coinciding headwinds. First of all, IP also suffered from weaker Chinese demand, as these customers have historically been significant IP licensees, so restrictions-driven caution has dragged on growth, with customers hesitant.Additionally and more importantly, the largest drag on growth was Synopsys’ deliberate, significant shift in resources away from lower-value, standalone IP blocks toward more complex, higher-value offerings such as subsystems and chiplet-ready IP optimized for die-to-die interconnects and advanced packaging.In standalone interface IP blocks, Synopsys could capture minimal long-term value, as the category is increasingly commoditized. Meanwhile, subsystems and chiplet IP are the opposite. They require deep integration expertise, foundry-specific optimization, and silicon validation that only a handful of vendors can credibly deliver. Synopsys is essentially moving up the value chain within its own IP business, accepting lower transaction volume in exchange for higher value per engagement and stronger customer stickiness.Ultimately, these higher-value subsystem and chiplet IP contracts mean larger deal sizes, better margins, and stronger customer lock-in. The royalty model layered on top means that as AI chip volumes scale and hyperscalers ship millions of units embedding Synopsys IP, revenue becomes increasingly volume-driven and self-compounding, with no proportional incremental cost.This is an upmarket move that improves Synopsys IP’s long-term position.However, it also created a significant short-term drag, as standalone blocks are relatively quick to license and deliver while subsystems and chiplet IP are more complex engagements. These require more customization, longer qualification cycles, deeper co-development with the customer, and, in many cases, tie revenue recognition to delivery milestones rather than upfront license fees. So, the sales cycle is significantly longer.In other words, Synopsys was essentially trading quick, smaller transactions for slower, larger ones, which creates a temporary growth gap as the portfolio transitions.Positively, IP revenue continues to grow strongly sequentially, with the bottom well behind it, while allowing Synopsys to fully benefit from growing hyperscaler demand. As AI drives demand for higher bandwidth, faster connectivity, and more complex system architectures, this drives strong growth for Synopsys’ interface, memory, and die-to-die IP. The company won more than 95% of PCIe 7 opportunities in the quarter and has secured 25 design wins YTD in LPDDR6, with the die-to-die business on pace to double this year, with more than 100 cumulative design wins – that is excellent momentum!I am really bullish on this improved IP positioning for Synopsys. Heightened design activity and a wave of new entrants building custom silicon, hyperscalers and AI labs alike, should keep expanding the pool of customers needing exactly this kind of high-value IP. And under the new structure, Synopsys is closing far larger deals, each one generating a royalty stream that compounds for years as those chips ship in volume. I expect revenue to keep accelerating in the coming quarters, and given current revenue growth (well ahead of expectations) and deal momentum, I am turning much more optimistic about its medium-term IP growth prospects.On that note, let’s move to the bottom-line results.Synopsys reported a Q3 GAAP gross margin of 72.5%, down 560 bps YoY, almost entirely reflecting a $200 million drag from amortization of acquired intangible assets. Really, it’s a non-cash accounting artifact of the Ansys deal, not a sign of deteriorating unit economics.Moving down the line, Synopsys reported non-GAAP operating expenses of $1.47 billion, on the lower end of the guided range, driven by operational efficiency and Ansys synergies running ahead of schedule. As a result, operating expenses were up just 21% YoY, with G&A down and R&D growth limited to 15%.This translated to an operating margin of 41.6%, up an excellent 310 bps YoY. As Synopsys realizes further synergies from the Ansys deal and improved top-line momentum enables more operating leverage, I expect Synopsys to keep delivering healthy margin expansion over the next several years.On a small separate note, SBC in Q3 came in at 9.4%, with SBC dollars down 5% YoY. SBC is still a bit elevated, partly due to the Ansys acquisition. We should see this trend down strongly over the next couple of years.Finally, non-GAAP EPS was $3.91, beating the consensus by $0.24, and Q3 FCF totaled $746 million, reflecting an excellent 30% FCF margin. Moreover, this brings the TTM total to a record $2.8 billion, reflecting a 29% FCF margin, as Synopsys’s margins are finally recovering.Under normalized conditions, this is an absolutely FCF machine, with margins likely to return to at least the mid-30s over the next couple of years.Driven by these strong cash flows, Synopsys maintained a solid balance sheet, ending the quarter with $3.6 billion in cash and roughly $11 billion in debt, for a net debt position of $7.2 billion. The Ansys acquisition last year moved Synopsys from a strong net cash position to net debt, but management has paid down this debt well over the last year, from $11.3 billion in net debt post-acquisition.Especially with cash flows improving strongly, I expect the balance sheet to keep improving. I certainly don’t see this as any sort of concern right now.Overall, I am really pleased with this quarter, with numbers across the board ahead of my prior expectations. Synopsys is showing broad-based strength and improving momentum in both EDA and IP amid strong demand dynamics. Meanwhile, margins and cash flows continue to recover, and the business remains in excellent financial health.I have little to complain about.Subscribe nowIs AI a Threat or an Opportunity?In late July, Moonshot AI (a Beijing-based lab backed by Alibaba) released Kimi K3, a 2.8-trillion-parameter open-weight MoE model that’s Moonshot’s largest and most capable coding/agentic model to date, positioned as competitive with Claude Fable 5 and GPT-5.6 Sol on several benchmarks.The issue for Synopsys? Moonshot claimed Kimi K3 autonomously designed a functional semiconductor chip in 48 hours using exclusively open-source EDA tools. According to Moonshot’s technical blog, K3 produced a 4mm² die running at 100MHz on the freely available Nangate 45nm Open Cell Library, completing an end-to-end RTL-to-tape-out-simulation flow without any licensed IP or proprietary software from Cadence or Synopsys, bypassing the giants entirely.Safe to say that caused quite the scare, with the EDA giants losing 9% of their value. Simply put, if an AI agent can orchestrate open-source EDA tooling well enough to tape out a chip, the moat protecting $10K+/seat licenses starts looking like it’s built on habit and switching costs rather than technical necessity. In other words, the news suggested that AI models could undercut Synopsys and Cadence.