
Synopsys
Recurring time-based software licenses + subscription/SaaS, silicon IP royalties + upfront, and (post-Ansys) simulation & analysis software — ~80%+ of revenue is recurring/time-based, sold to semiconductor and systems companies.
Sources — 7 figures with citations
- Latest print: Q3 FY2026 (quarter ended 2026-07-31), reported 2026-08-26filed2026-11-12revenue $2,476.8M; about +42% YoY (vs $1,740M; includes Ansys); guide: Q4 FY2026 revenue $2,530M-$2,580M, non-GAAP EPS $4.10-$4.16; FY2026 revenue $9,690M-$9,740M, non-GAAP EPS $15.04-$15.10news.synopsys.com — Recorded Nov 12 2026; figures as the source states them. The entry's analysis text predates this print unless dated otherwise.
- earnings, 2026-08-26filed2026-08-26Q3 FY2026 results (quarter ended Jul 31, 2026): revenue $2,476.8M vs $1,740M a year earlier (about +42% YoY, Ansys included); GAAP EPS $2.84 (vs $1.50) / non-GAAP $3.91 (vs $3.39); Design Automation $2,003.0M (80.9%), Design IP $473.8M (19.1%, back to YoY growth from $427.6M); restructuring charges about $2.2M in the quarter and $236.3M for the nine months.news.synopsys.com — CONFIRMED; recorded Nov 12 2026; source dated 2026-08-26
- guidance, 2026-08-26filed2026-08-26Guidance raised: FY2026 revenue $9,690M-$9,740M (midpoint $9.715B, from $9.665B) including $2.98B of expected Ansys revenue; non-GAAP EPS $15.04-$15.10 (from a $14.76 midpoint); GAAP EPS $3.84-$4.08; Q4 FY2026 revenue $2,530M-$2,580M, non-GAAP EPS $4.10-$4.16, GAAP EPS $0.60-$0.85; targets 'assume no further changes to export control restrictions or the current U.S. government Entity List restrictions'.news.synopsys.com — CONFIRMED; recorded Nov 12 2026; source dated 2026-08-26
- other, 2026-10-26filed2026-10-26Pair peer Cadence Q3 2026: revenue $1.633B vs $1.339B (+22% YoY), record backlog $8.2B ($4.2B to convert within 12 months), FY2026 revenue guide raised a third time to $6.39B-$6.42B (about 21% growth) with non-GAAP EPS $8.37-$8.43, a $2.5B buyback added, China 16% of revenue; guidance assumes no change in export controls.sec.gov — REPORTED; recorded Nov 12 2026; source dated 2026-10-26
- Watched signal: FY26 guidance raises (revenue to $9.665B, non-GAAP EPS to $14.76)filed2026-11-12Raised again on Aug 26 to $9.69B-$9.74B revenue and $15.04-$15.10 non-GAAP EPS.news.synopsys.com — Outcome of a signal this entry said to watch; recorded Nov 12 2026.
- Watched signal: Design IP segment growth (fell ~6% YoY Q1 FY26)filed2026-11-12Returned to growth: $473.8M in Q3 FY2026 vs $427.6M a year earlier (about +11%).news.synopsys.com — Outcome of a signal this entry said to watch; recorded Nov 12 2026.
- Watched signal: Cadence relative performance (+19% YoY, record $8B backlog)filed2026-11-12Cadence grew 22% YoY to $1.633B in Q3 2026 with a record $8.2B backlog and FY2026 growth guidance of about 21%; Synopsys printed about +42% reported (Ansys-inflated) with no organic rate disclosed, so the organic gap is still unmeasured.sec.gov — Outcome of a signal this entry said to watch; recorded Nov 12 2026.
The thesis on this name
State of AI Compute
Widest, longest-half-life EDA moat (foundry-certified flow, ~2/3 of 3nm exclusive, AI-design corpus) at the cheapest multiple of the three toll-keepers — but it is not cheap on EV/EBITDA (57-67x live, not ~22x), so the LONG shrinks toward a relative-value pair vs CDNS, not an outright cheap-compounder.
State of AI Compute
Long Synopsys over a 12-18mo horizon as a RELATIVE-VALUE pair vs Cadence: SNPS owns the widest, longest-half-life EDA moat (foundry-certified flow, ~2/3 of leading-edge 3nm tape-outs) yet trades at a persistent ~8-10x forward-P/E and ~12-17x EV/EBITDA discount to CDNS that is unjustified by the fundamental gap — the edge is the analytical mis-pricing of the spread, not absolute cheapness (SNPS is ~33-40x EV/EBITDA, expensive outright).
State of AI Compute
SNPS verdict was avoid as an outright, but the CDNS verdict says express the real EDA edge as long-SNPS / short-CDNS to harvest the duopoly relative-value gap while neutralizing shared AI-capex beta. Token, catalyst-dated into the Aug Q3 print with a tight +1.5 P/E-turn spread stop — not the 12-18mo structural convergence bet.
Earnings, margins, COGS & capex
Synopsys is in a transition year: headline revenue is up ~40%+ YoY but the jump is the July-2025 Ansys acquisition, while organic EDA grows low-double-digits and Design IP is soft (China + a transitional year). Software gross margins sit in the low-70s% and non-GAAP segment operating margin is ~39-40% with a stated path to 40%+ as Ansys synergies (~$400M cost run-rate by yr3) land. GAAP earnings are temporarily crushed by Ansys purchase-accounting amortization (GAAP EPS $0.09 vs non-GAAP $3.35 in Q2 FY26); the real story is recurring revenue durability plus margin expansion. Net debt of ~$7.5B from the deal is the main balance-sheet change from a historically net-cash company.
Income statement — where each revenue dollar goes
% of revenueOf every $1 of revenue, ~28¢ is cost of goods and ~67¢ operating expense, leaving ~5¢ of operating profit (~28¢ net).
Revenue trend
Margins
flat/slightly down — Ansys simulation mix and IP royalty timing dilute slightly vs pure-EDA ~80%+ (fact)
up — cost discipline + Ansys synergies, management targets 40%+ (fact)
down sharply — Ansys intangible amortization + integration; normalizes over time (fact)
stable (fact)
down vs historical mid-30s% — integration cash costs + higher interest on deal debt (fact/estimate)
COGS structure
Software-business COGS is dominated by (1) capitalized-software and acquired-technology amortization — now greatly inflated by Ansys purchase accounting — (2) hosting/cloud compute for cloud-EDA and AI workloads, (3) silicon-IP productization/qualification cost (test chips, foundry certification), and (4) field application / technical support headcount. There is essentially no physical-goods COGS; the 'cost' is engineering amortization and support, which is why gross margins are structurally low-70s-to-80s%.
Capex
Minimal and asset-light: ~$89.5M for 1H FY26 (~1.5-2% of revenue). Funds R&D lab/compute infrastructure, data-center/cloud capacity for EDA-on-cloud and AI-assisted design, and office build-out — not factories. The real 'investment' runs through opex R&D, not capex.
Latest earnings
Beat — revenue $2.276-2.28B vs ~$2.25B consensus; non-GAAP EPS $3.35 vs ~$3.15 consensus (+$0.20) (fact).
Raised FY26: revenue to $9.665B midpoint, non-GAAP EPS to $14.76 midpoint, FCF ~$2.0B; Q3 FY26 revenue $2.41-2.46B, non-GAAP EPS $3.63-3.69 (fact).
- Design Automation revenue (Q2 FY26)
- $1.822B (80% of total)
- Design IP revenue (Q2 FY26)
- $454.2M (20%; soft — IP fell ~6% YoY in Q1 FY26)
- Non-GAAP net income (Q2 FY26)
- $643.7M
- China revenue
- ~9% of total in Q1 FY26 (vs ~16% FY24) — still depressed post-2025 controls
Growth drivers
- AI-compute design boom — every AI accelerator, HBM-attached SoC, and chiplet design pulls more EDA tool seats, IP, and compute-heavy verification (fact-led, secular).
- Rising design complexity at 3nm/2nm and below + chiplets/3D-IC — more transistors, more verification cycles, more tool consumption per design (structural).
- Ansys cross-sell — multiphysics simulation attached to the chip-to-system flow; management targets ~$400M run-rate revenue synergy by yr4, >$1B long-term (estimate, company-stated).
- AI-native EDA (Synopsys.ai / generative + agentic design) lifting price/seat and stickiness as autonomous design closes the engineer-shortage gap (fact/estimate).
- Silicon IP secular growth — outsourcing of interface/foundation IP (PCIe, HBM, UCIe, USB) as design teams buy rather than build, once China and the transitional-year drag fade (mixed/estimate).
- China revenue normalization — a recovery from the 2025 export-control shock (China ~9% of revenue in Q1 FY26 vs ~16% in FY24) is upside optionality, not base case (estimate).
Reported financials — SEC EDGAR
Audited GAAP figures pulled from SEC filings · latest filing 2025-12-22. The audited primary-source spine — not financial advice.
Revenue — annual (GAAP)
Margins & balance sheet — FY’25
Bull & bear
Own the widest, most durable moat in semiconductors at a persistent discount to its near-twin Cadence — a foundry-certified EDA franchise compounding on the AI-design supercycle, with Ansys synergies and a China-recovery option as un-paid-for upside, set to re-rate as integration de-risks.
- Moat + relative value: #1 EDA share (~31%), foundry-certified across a majority of 3nm tape-outs, yet trades ~25x NTM EV/EBITDA vs Cadence ~34x — an ~8-10x forward-P/E and meaningful EV/EBITDA discount to a slower-moat-half-life peer.
- Secular demand: AI-compute, 3nm/2nm complexity, and chiplets structurally raise EDA tool + IP consumption per design — Synopsys sells the picks-and-shovels to the entire chip industry regardless of which chip wins.
- Margin expansion catalyst: management targets 40%+ non-GAAP operating margin via cost discipline + ~$400M Ansys cost-synergy run-rate; raised FY26 EPS to $14.76 shows the lever is already turning.
- Vertically integrated, un-replicable bundle: only player with EDA + IP + Ansys multiphysics — a silicon-to-system stack that deepens lock-in and opens a >$1B long-term revenue-synergy pool.
- Asymmetric China option: China at ~9% of revenue (vs ~16% pre-controls) means the bad news is largely in the price; any normalization is upside the current multiple isn't paying for.
You're paying a high-30s P/E for a company whose headline growth is an acquisition, whose GAAP earnings are near-zero, that just took on ~$10B of debt, and whose faster-growing twin Cadence is winning the AI-design narrative — while a structural China/export-control overhang can re-fire at any time.
- Growth is bought, not organic — the +40%+ YoY is Ansys; organic EDA is low-double-digit and Design IP is shrinking (~-6% YoY), so the multiple rests on a synergy story that isn't yet proven.
- Cadence is out-executing: +19% YoY organic, record $8B backlog, premium multiple — the relative-value 'discount' may persist because the market is rationally paying up for the faster, cleaner grower.
- Balance-sheet and quality dent: ~$7.5B net debt and a GAAP EPS of $0.09 vs $3.35 non-GAAP mean reported profitability is heavily adjusted; integration cash costs pressure near-term FCF.
- Structural China risk — export-control whiplash already vaporized ~half of China revenue once and accelerated domestic Chinese EDA substitution; the option works both ways and could re-fire.
- Valuation leaves little margin of safety — a high-30s forward P/E on a transition-year, debt-laden integration prices in flawless synergy delivery and a benign policy/cycle backdrop.
What it is worth
Relative-value pair vs Cadence (forward EV/EBITDA + P/E), cross-checked against SNPS's own normalized FCF
~$380-420 (back toward the 52-wk low ~$376 on synergy slippage, a China re-tightening, or a semi-cycle/multiple de-rate)
~$520-560 (≈34-37x FY27E EPS ~$15.5-16.5; partial discount compression as integration de-risks)
~$600+ (≈40-42x FY27E non-GAAP EPS / EV/EBITDA re-rating toward Cadence as Ansys synergies land + China normalizes; approaches the prior 52-wk high of $651.73)
Core thesis: SNPS trades ~25x NTM EV/EBITDA / high-30s forward P/E vs Cadence ~34x — own the wider-moat name into synergy-driven re-rating; base case is partial gap-closure, not full parity.
SWOT
Strengths
- Widest, longest-half-life EDA moat — foundry-certified digital + custom flows used on a majority of leading-edge (3nm) tape-outs; #1 EDA share (~31%, fact).
- Recurring, time-based revenue model (~80%+ recurring) with multi-year contracts and very high renewal — revenue is sticky and visible.
- Now the only vertically integrated silicon-to-system stack (EDA + IP + Ansys multiphysics) — a differentiated bundle Cadence/Siemens can't fully match.
- Structural beneficiary of AI-compute design demand and rising 3nm/2nm complexity — both lift tool/IP consumption per design.
- Strong software economics — low-70s%-to-80s% gross margin, ~40% non-GAAP operating margin, ~30% FCF margin, minimal capex.
Weaknesses
- Net debt of ~$7.5B post-Ansys (from a historically net-cash company) plus higher interest expense — financial flexibility reduced during integration.
- GAAP earnings temporarily crushed by Ansys purchase-accounting amortization (GAAP EPS $0.09 vs non-GAAP $3.35) — large GAAP/non-GAAP gap dents quality perception.
- Design IP segment is soft and lower-margin in a 'transitional year' — IP fell ~6% YoY in Q1 FY26.
- Heavy reliance on a handful of advanced-node customers (TSMC ecosystem, hyperscalers, large fabless) — concentrated end-demand.
- Integration execution risk — merging Ansys's go-to-market, channel accounting, and ~$400M synergy delivery is multi-year and not yet de-risked.
Opportunities
- Ansys cross-sell into the chip-to-system flow — company targets >$1B long-term revenue synergy plus ~$400M cost synergy run-rate.
- AI-native / agentic EDA (Synopsys.ai) raising price-per-seat, stickiness, and TAM as autonomous design offsets the engineer shortage.
- China revenue recovery optionality — a normalization from ~9% back toward the historical ~16% would be material upside (currently not in the base case).
- 3D-IC / chiplet / UCIe design and verification — a rising multi-die TAM where Synopsys IP + tools + Ansys thermal/structural simulation all attach.
- Re-rating toward Cadence's multiple as integration de-risks — SNPS trades ~25x NTM EV/EBITDA vs CDNS ~34x, a persistent gap that can compress.
Threats
- U.S.-China export-control whiplash — the May-2025 EDA ban (rescinded July 2025) showed China revenue (~16% in FY24) can be switched off overnight; the policy risk is structural, not one-time.
- Cadence is a near-equal, faster-organic-growth competitor (CDNS +19% YoY Q1 2026, record $8B backlog) gaining share in AI-design narrative and trading at a premium.
- Chinese domestic EDA substitution (Empyrean, others) accelerated by sanctions — a long-tail share threat in the world's largest chip-build market.
- Customer-base concentration + semiconductor cyclicality — a capex pullback by foundries/hyperscalers compresses tool and IP demand.
- Higher-for-longer rates raise the interest cost on the ~$10B of acquisition debt and pressure the multiple of a long-duration growth name.
Moats, dependencies & bottlenecks
Moats
Very high (10yr+) Synopsys flows are co-developed and certified with TSMC/Samsung/Intel for each new node; switching tools mid-design risks tape-out — the deepest source of lock-in.
Designs, scripts, and IP are built around Synopsys tooling; an entire generation of design engineers is trained on it — replacing it means re-validating a working flow.
#1/#2 in commercial IP (PCIe, HBM, UCIe, USB, foundation); buying proven, foundry-qualified IP is faster and safer than building — reinforces the tool relationship.
Medium-High (proving) Only vendor pairing EDA + IP + Ansys multiphysics; a real differentiator but the cross-sell synergy is still being demonstrated, so not yet 'wide'.
Generative/agentic design tools compound the data/scale advantage, but Cadence is investing equally hard — an arms race, not a settled moat.
Dependencies
Tools must be certified for each foundry's new node; loss of a certification partnership or a foundry roadmap stall directly hits tool relevance.
May-2025 EDA ban (rescinded Jul-2025) showed China (~16% of FY24 revenue, ~9% now) can be cut off by policy overnight — recurring binary overhang.
Tool and IP demand tracks new-design starts; a foundry/hyperscaler capex pullback compresses consumption even with sticky base licenses.
A handful of advanced-node customers (Nvidia, AMD, Apple, hyperscaler silicon teams) drive a disproportionate share of leading-edge tool/IP spend.
FY26 guidance and the re-rating thesis assume ~$400M cost + >$1B long-term revenue synergies land; slippage is the key controllable execution risk.
Advantages
- Largest EDA franchise (~31% share) with foundry-certified flows on a majority of leading-edge 3nm tape-outs — the default tool for the most advanced chips.
- Recurring, time-based revenue (~80%+) with high renewal and multi-year visibility — durable, predictable cash generation.
- Only vertically integrated silicon-to-system stack (EDA + IP + Ansys) — a bundle that deepens lock-in and expands wallet share.
- Software economics — low-70s%-to-80s% gross margin, ~40% non-GAAP operating margin, ~30% FCF margin, minimal capex.
- Direct, diversified leverage to the AI-compute and advanced-node supercycle — sells to the whole industry, agnostic to which chip wins.
Weaknesses
- Headline growth is acquisition-driven — organic EDA is low-double-digit and Design IP is shrinking (~-6% YoY) — slower organic than Cadence.
- ~$7.5B net debt post-Ansys (from net-cash) plus interest expense — reduced flexibility during integration.
- Large GAAP/non-GAAP gap (GAAP EPS $0.09 vs $3.35 non-GAAP) from purchase-accounting amortization dents earnings quality optics.
- Structural China/export-control exposure that can be switched off by policy — and is accelerating domestic Chinese EDA substitution.
- Design IP segment is lower-margin and in a 'transitional year' — a drag on consolidated margin and growth.
- High-30s forward P/E leaves little margin of safety if synergy delivery or the semi cycle disappoints.
Bottlenecks
- Ansys integration capacity — finite management/engineering bandwidth to merge channels, products, and back-office while hitting synergy targets on schedule.
- Export-control / licensing gates — China access is governed by BIS policy the company doesn't control; a binary cap on ~10-16% of TAM.
- Foundry node cadence — tool revenue per new node is paced by how fast TSMC/Samsung/Intel advance to 2nm/A16; a foundry roadmap slip slows the upgrade cycle.
- Skilled-engineer scarcity in both customers and Synopsys — the same chip-design talent shortage that creates AI-EDA demand also limits how fast new tools/IP are built and adopted.
- Debt-service drag — ~$10B acquisition debt and higher rates cap buyback/M&A optionality until leverage normalizes.
Top signals & trends
Top signals
7 sourced items since this entry's vintage are in the evidence rows (latest print, guidance, deals, resolved watch items); the analysis text is as of the entry's stated date and predates them unless dated otherwise.
Q2 raise on cost discipline + accelerating Ansys synergies validates the margin-expansion leg of the thesis.
Watch the trajectory — re-normalization is upside optionality; another export-control tightening is the key downside binary.
IP softness is the clearest organic-weakness tell; a return to growth would de-risk the 'transitional year' narrative.
The pair trade's risk — if Cadence keeps out-growing, the SNPS discount can persist or widen rather than close.
The core relative-value setup; compression as integration de-risks is the re-rating catalyst.
On-track synergy disclosure each quarter is the proof point the bull case hinges on.
Trends
Every AI accelerator, HBM stack, and chiplet pulls more EDA tools, IP, and verification compute — secular tailwind for the whole Big-3.
More transistors and multi-die integration mean more tool consumption and more multiphysics simulation per design — favors the integrated stack.
Caps the China TAM and seeds domestic Chinese EDA rivals, but also entrenches Synopsys with allied foundries; a recurring binary risk.
Generative + agentic design lifts price-per-seat and stickiness and offsets the engineer shortage — but it's an arms race with Cadence.
The Ansys deal and Cadence's systems push show the market rewarding integrated multiphysics + EDA bundles over point tools.
Sticky recurring licenses cushion downturns, but new-design and IP demand still flex with the foundry/hyperscaler capex cycle.
Ecosystem & competitor graph
Suppliers feed the company; customers pull from it. Line thickness shows the strength of each tie (supply-chain dependency, customer earnings contribution). Hover to isolate a tie.
Cloud-compute providers running EDA-on-cloud and AI-assisted design workloads — a key COGS and capability input.
Cloud infrastructure partner/supplier for hosted EDA and verification-at-scale.
Licenses CPU architectures that Synopsys IP and tools must support/interoperate with — an upstream IP input.
GPU supplier for AI-EDA / accelerated verification and an EDA customer — a dual supplier-customer relationship.
Co-develops and certifies Synopsys flows per node; both the top foundry partner and the gateway to fabless customers.
Leading AI-chip designer — a marquee advanced-node EDA + IP customer; also an AI-EDA partner.
CPU/GPU/accelerator designer; heavy consumer of leading-edge digital + custom flows and IP.
In-house silicon (A/M-series) on advanced nodes — large tool + IP consumer.
Both a foundry partner (certification) and a chip-design customer for tools and IP.
Custom-silicon / ASIC leader for hyperscalers — major EDA and interface-IP customer.
Mobile/compute SoC designer — long-standing tools + IP customer.
Near-equal EDA share (~30%), faster organic growth (+19% YoY Q1 2026, record $8B backlog), premium multiple — the direct pair-trade counterpart.
#3 in EDA (~13% share), strong in verification/PCB/IC test; backed by Siemens' scale but not a leading-edge digital-flow threat at SNPS's level.
Former standalone multiphysics leader — acquired Jul-2025; now an asset, not a competitor.
Overlaps in design/simulation and EM/electronic test (PathWave); adjacent rather than head-to-head in core IC EDA.
Chinese EDA vendors (Empyrean, others) gaining state-backed momentum as sanctions push China toward home-grown tools — a long-tail share threat in the largest chip-build market. Named for analysis only; not a recommendation.
Competes/partners in the processor-IP layer; an IP rival in CPU cores but also an ecosystem partner — adjacent to core EDA tools.