15 EDA Silicon IP
The EDA and silicon IP sector is a structural beneficiary of rising semiconductor complexity, driven by AI chip proliferation, migration to advanced nodes (3nm/2nm), and demand for pre-verified building blocks (CPU cores, memory interfaces, security IP). With the market projected to grow at 6–14% CAGR through 2035, the thesis holds that pure-play IP vendors and EDA duopolists will compound value through recurring license royalty streams and rising R&D spend by fabless design houses. However, extreme valuation dispersion across the group demands careful selectivity — the structural tailwind is real but already priced into some names at levels that leave no room for execution error.
| Outlook | bull |
|---|
| Sentyment sektora | Bullish |
|---|
| Faza cyklu kapitałowego | mid_cycle |
|---|
| Ekspozycja na AI | Very high. AI chip design (GPU, ASIC, custom accelerators) is the primary demand driver for advanced EDA flows and processor/memory IP. Every new AI SoC requires multiple IP blocks from this group. |
|---|
Uzasadnienie outlooku
SoC complexity, node migration to 3nm/2nm, and the secular shift toward AI-first computing architectures are expanding TAM for both EDA tools and silicon IP. The 2026 wave of M&A (Synopsys/Ansys, Qualcomm/Alphawave, GlobalFoundries/ARC) confirms strategic value of IP. Despite near-term macro uncertainty, the structural demand curve remains intact.
Ranking spółek
- SNPS — Synopsys ()
- CDNS — Cadence ()
- ARM — ARM Holdings ()
- RMBS — Rambus ()
- 3529.TWO — eMemory Technology ()
Kluczowe ryzyka
- {'risk': 'AI Capex Slowdown: If hyperscaler AI investments peak in 2026, tape-out volumes could decline. SNPS/CDNS have recurring revenue buffers; ARM and eMemory are more exposed to licensing cycles.', 'timeframe': '12–18 months'}
- {'risk': "Geopolitical Fragmentation: US-China export controls may create two distinct design ecosystems (US-aligned vs. China-autarky). This could fragment the IP market, erode ARM's royalty base in China, and force EDA tool customization, raising costs.", 'timeframe': '24 months'}
- {'risk': "RISC-V Threat: Increasing adoption of open-source RISC-V in edge AI and automotive could erode ARM's CPU IP moat over time. Synopsys and Cadence support RISC-V, so they are partially hedged.", 'timeframe': '36 months'}
- {'risk': 'Valuation Multiple Compression: Current forward P/E multiples (SNPS 25.8x, CDNS 40.9x, ARM 105x) are not cheap in absolute terms. Rising interest rates or an economic downturn could re-rate these names lower by 20–30%.', 'timeframe': '6 months'}
- {'risk': "Supply Chain Concentration (eMemory-specific): Taiwan's geopolitical risk makes eMemory's revenue base vulnerable to disruption. Any credible escalation in cross-strait tensions could trigger a permanent valuation discount.", 'timeframe': '12 months'}
Katalizatory
- {'catalyst': "Q2 2026 Earnings (eMemory, SNPS, CDNS): Positive revisions and AI licensing wins would affirm the growth trajectory. eMemory's Q1 2026 already showed acceleration (+20% revenue).", 'timeframe': 'late July / early August 2026'}
- {'catalyst': 'ARM Neoverse CCS V3 Launch: Next-gen server cores with higher performance would strengthen the data center growth narrative.', 'timeframe': '2026 H2'}
- {'catalyst': 'NSIT / NATO Security Certification for NeoPUF (eMemory): Would open government and secure enclave TAM, boosting royalty visibility for eMemory.', 'timeframe': '2026 H2'}
- {'catalyst': 'GlobalFoundries / Synopsys ARC IP Deal Closing: Adds to the narrative that IP is a strategic acquisition target, potentially re-rating the entire subsector.', 'timeframe': '2026'}
- {'catalyst': 'Continued M&A in IP Space: Further consolidation (e.g., Arm acquiring an IP company) could crystallize scarcity value for pure-play IP assets.', 'timeframe': 'medium-term'}