Snapshot

The global analog and mixed-signal semiconductor market generates roughly $80–90B in annual revenue est., of which five major US-listed companies — TXN, ADI, NXPI, ON, and MCHP — collectively produce about $55B in trailing-twelve-month sales and carry a combined enterprise value of ~$700B. Analog chips convert real-world signals (voltage, temperature, pressure) to digital data and back; mixed-signal chips combine both functions on one die. These components sit in every electronic system — cars, factories, phones, data centers — but individually cost cents to single-digit dollars, making the revenue stream broad-based and tied to unit volumes across the entire economy. The sub-sector is emerging from a deep inventory correction that bottomed in mid-2024, with most companies now reporting sequential quarterly revenue growth for 3–4 consecutive quarters.

~$700B in combined enterprise value buys ~$55B of revenue and ~$10B of free cash flow. Revenue peaked in 2022 at ~$62B for these five and is recovering from a ~15% cycle-driven trough. Analog chips are not on the critical path for AI compute itself, but they are ubiquitous in every end market that AI is accelerating — data center power delivery, automotive electrification, and industrial automation.

The product & how money is made

Analog chips handle real-world, continuous signals — voltage regulation, signal amplification, sensor reading, motor control. Mixed-signal chips bridge both domains: they contain analog-to-digital converters (ADCs) and digital-to-analog converters (DACs) on the same die. A single modern car contains $50–80 worth of analog/mixed-signal content est. (power management ICs, sensor interfaces, motor drivers, communication transceivers); an AI server rack contains $10–30 worth est. (voltage regulators, data converters, power-stage controllers).

Revenue mix by end market

• Automotive (30–40% of sub-sector revenue): ADAS sensors, battery management for EVs, body electronics, infotainment. NXPI (~50% auto revenue) and ON (~50% auto/industrial) are most exposed. TXN derives ~25% from auto.
• Industrial (25–35%): factory automation, building controls, grid infrastructure, test and measurement. TXN's largest segment. Cycle-sensitive: follows capital spending patterns with 6–12 month lags.
• Communications infrastructure (10–15%): base stations, data center power delivery, high-speed data converters for optical/wired links. ADI has the highest exposure (~20%).
• Consumer and personal electronics (10–15%): smartphones, wearables, IoT devices. Shortest product life cycles and most price-sensitive.
• Enterprise/computing (5–10%): server power management, storage controllers. Smallest direct contribution but fastest recent growth driver via AI data center buildout.

How the money flows

Design a chip once (R&D cost: $1–10M per product, 2–4 year development cycle), then sell it for 10–20 years with minimal redesign. Analog designs change slowly because the underlying physics doesn't change — a voltage regulator designed in 2015 may still be the best option in 2026. Gross margins run 55–65% for established product lines because wafer costs are low (mature nodes) and the design IP is the value. TXN and ADI each have catalogs of 80,000+ products; no single product exceeds ~1% of revenue.

Manufacturing split: TXN is fully vertically integrated — it owns and operates its own fabs (including a massive expansion program). ADI, NXPI, and MCHP are fab-lite — they own some older fabs but outsource to TSMC, GlobalFoundries, and others for newer designs. ON owns SiC (silicon carbide) fabs for power semiconductors but outsources standard analog.

Sources: TXN 10-K FY2025, ADI 10-K FY2025, NXPI 20-F FY2025, ON 10-K FY2025, MCHP 10-K FY2026 (Mar fiscal year-end).

Demand

Contracted / visible demand

Analog semiconductor companies generally do not disclose large multi-year contracted backlogs. Order patterns are shorter-cycle: customers place orders 4–16 weeks ahead (lead times normalized from 52+ weeks during the 2021–2022 shortage). Exceptions:

• Automotive platform wins are effectively contracted: once an OEM qualifies a chip into a vehicle platform (18–24 month qualification), that design stays for the 5–7 year production life. These wins are not disclosed as dollar-value backlogs but represent high-visibility multi-year revenue streams.
• TXN's CHIPS Act grants — up to $1.6B from the US government (announced 2024) to support 300mm fab construction in Texas and Utah — represent contracted funding, not revenue, but underpin capacity expansion.
• Long-term supply agreements (LTSAs): during the 2021–2022 shortage, customers signed 1–3 year LTSAs guaranteeing minimum purchase volumes. Most have expired or been renegotiated down, contributing to the 2023–2024 inventory correction.

Demand forecasts

The analog market is cyclical, driven by global industrial production, auto builds, and consumer electronics shipments. est.

• Automotive electrification: an EV uses roughly 2× the analog content of an ICE vehicle ($80–100 vs $40–50) est., driven by battery management, onboard chargers, and traction inverters. Global EV penetration was ~20% of new car sales in 2025 and is projected to reach 35–45% by 2030. est.
• AI data center power delivery: each AI server rack draws 40–120 kW and requires dozens of voltage regulator modules (VRMs), current sensors, and power-stage controllers. Analog content per AI rack is $10–30 est., versus $200,000–500,000 of GPU content — a growth catalyst at the margin, not a revenue transformation.
• Industrial automation and grid modernization: steady mid-single-digit secular growers. est.
• Near-term recovery: after a ~15% peak-to-trough decline (2022 peak to 2024 trough), the sub-sector is in early recovery. TXN Q1 2026: $4.83B revenue, up 22% YoY. ADI Q2 FY2026 (Apr 2026): $3.62B, up 26% YoY.

Sources: TXN Q1-2026 10-Q, ADI Q2 FY2026 10-Q, MCHP Q4 FY2026 10-Q, ON Q1-2026 10-Q, NXPI Q1-2026 20-F. Market size estimates from WSTS and IC Insights; EV penetration from IEA Global EV Outlook 2025.

Supply

Capacity and manufacturing

Analog chips are manufactured on mature process nodes (28nm and older, primarily 65nm–180nm). Unlike leading-edge digital chips, analog designs gain little from shrinking transistors — performance depends on precision, noise characteristics, and voltage handling rather than transistor density.

• No structural shortage. The 2021–2023 analog chip shortage was demand-driven (pandemic electronics + auto restocking), not capacity-limited. Capacity has fully caught up; inventory levels at customers and distributors are back to or above normal.
• TXN's fab buildout: $5B/year capex (2023–2025: cumulative ~$14.4B) to build four new 300mm fabs — RFAB2 (operational), LFAB (ramping), and two in Sherman, TX (under construction). When fully equipped, these roughly double TXN's wafer output. est. Net PP&E grew from $6.9B (2022) to $12.3B (2025). FCF compressed from ~$6.5B (2023, pre-ramp) to $1.1B trailing.
• ON's SiC capacity: SiC fabrication in Hudson, NH and substrate facility in South Korea. SiC chips command 3–5× the ASP of silicon equivalents est.. ON reduced total capex to $341M in 2025 (from $1.5B in 2023) as EV growth slowed.
• ADI, NXPI, MCHP are predominantly fab-lite: FY2025 capex was $534M (ADI), $537M (NXPI), $91M (MCHP). They rely on foundry partners for wafer production.

Bottlenecks

The binding constraint is demand, not fabrication capacity. Unlike GPUs or HBM, there is no waitlist for analog chips. Supply-side risks:

• Geopolitical: ~60% of analog wafer production occurs in Asia (Taiwan, China, South Korea). est. TXN's domestic fab buildout is partly a hedge against this concentration.
• Inventory overhang: as of Q1 2026, TXN had $4.8B of inventory (167 days of sales), well above the historical ~120-day norm.
• SiC substrate supply: ON, STMicroelectronics, and Wolfspeed compete for limited SiC wafer substrates. A niche bottleneck specific to high-power automotive.

Sources: TXN 10-K FY2025 (capex, fab status, inventory); ON 10-K FY2025 (SiC operations); ADI, NXPI, MCHP annual filings.

The gap

The analog/mixed-signal sub-sector is in oversupply relative to current demand, with supply growing (especially at TXN) while demand recovers from a cyclical trough — the opposite dynamic from AI compute (GPUs, HBM).

Pricing direction: analog chip pricing is flat to slightly declining. During the 2021–2022 shortage, TXN and ADI raised prices 5–15% est., but those increases have partially reversed. Revenue growth comes from volume (more chips per system, more systems) rather than pricing power.

Margin context: TXN gross margin was 69% at 2022 peak vs 57% trailing. ADI was 64% at peak vs 65% trailing (already recovered). ON was 49% at peak vs 43% trailing (still depressed, partly SiC-related).

Sources: Company 10-Q/10-K filings (revenue, inventory, margins); lead time data from SIA and company earnings calls.

The players

Company-specific notes

• TXN — largest analog company by revenue. Unique among peers in owning 300mm fabs outright. Currently in its heaviest capex cycle ($5B/yr), depressing FCF from a normalized ~$6B+ to ~$1B. TXN claims 300mm wafers cost 40% less per chip than 200mm. Returned $6.5B to shareholders in 2025 ($5.0B dividends, $1.5B buybacks), funded partly by debt — total debt rose from $8.7B (2022) to $14.0B (2025). Dividend yield 1.8%.
• ADI — highest gross margins (65%), reflecting premium positioning in high-performance data converters, RF, and precision signal chain. Acquired Maxim Integrated in 2021 for $21B ($34B goodwill/intangibles; tangible book value -$487M). Latest quarter $3.62B, up 26% YoY. Returned $4.1B to shareholders in FY2025 ($2.2B buybacks + $1.9B dividends).
• NXPI — most automotive-exposed (~50% of revenue). Dutch-incorporated, files 20-F. Revenue flattish ($12.3–13.3B) for four years. Cheapest on EV/Revenue (7.1×) and highest FCF yield (3.3%). Carries $11.7B debt against $3.7B cash. Tangible book value negative (-$858M) due to $11.8B of acquisition-related intangibles.
• ON — the SiC/power semiconductor play. Revenue declined 28% from 2022 peak ($8.3B to $6.0B in 2025). Gross margins compressed from 49% peak to 43%. No dividend; returned $1.4B via buybacks in 2025. Posted a net loss in Q1 2026 (-$33M).
• MCHP — primarily a microcontroller company with analog/mixed-signal as secondary (~35% of revenue). Revenue fell 44% from FY2023 peak ($8.4B) to FY2026 ($4.7B). Net income went from $2.2B to $230M. Carries $5.5B debt against $240M cash. Q4 FY2026 revenue of $1.31B was up 22% sequentially.

Sources: yfinance (prices, market cap as of Jun 2 2026); TXN/ADI/ON/MCHP 10-K and 10-Q; NXPI 20-F FY2025.

The price of exposure

Implied math

• TXN at $308: $280B market cap for $5.0B net income and $1.1B FCF (depressed by $4.6B/yr capex). If capex normalizes to ~$2B/yr after the fab build completes (~2028) est., normalized FCF would be ~$5–6B — implying ~2% normalized FCF yield. At peak 2022 earnings ($8.7B net income, 69% gross margin), trailing P/E would have been ~32×.
• ADI at $423: $206B market cap for $3.9B FCF. Margins are near peak and capex is modest — ~2% FCF yield at face value. $34B of goodwill from the Maxim deal means 70% of the balance sheet is intangible.
• NXPI at $324: cheapest on EV/Revenue (7.1×) and highest current FCF yield (3.3%). If revenue recovers to 2022–2023 levels (~$13.2B), current EV/Revenue drops to ~6.8×. $11.7B debt is 4.3× trailing FCF.
• ON at $129: at 52-week high despite a net loss last quarter. If margins recover to 2023 levels (47% gross, $2.2B net income), trailing P/E at today's price would be ~23×.
• MCHP at $97: revenue 44% below peak; trailing P/E of 441× reflects trough earnings. At FY2024 peak earnings ($1.9B net income), today's price implies ~28× peak P/E. $5.4B net debt is 23× trailing FCF.

Sources: yfinance live data (Jun 2 2026); company filings for historical earnings and capex.

What to deep-dive next

• TXN's fab economics: the $14.4B capex program is the biggest bet in the sub-sector. Key question: incremental return on invested capital once fabs are loaded. Verify unit economics from 10-K disclosures and CHIPS Act filings.
• NXPI's automotive moat: at 7.1× EV/Revenue and 3.3% FCF yield, NXPI is the cheapest name here. Determine whether flat automotive revenue (4 years) is secular or cyclical. Pull the 20-F for design-win pipeline and content-per-vehicle trajectory.
• ON's SiC position: gross margins compressed 6 points while peers recovered. Determine whether this is temporary (EV destocking) or structural (SiC competition from STMicro, Wolfspeed, Infineon).
• Inventory normalization timing: TXN's 167-day inventory vs 120-day historical norm is the clearest cycle-turn indicator. Track quarter-over-quarter drawdown.
• AI power-delivery content: quantify actual analog dollar content per AI server rack across the five companies — frequently cited as a growth catalyst but may be $10–30 per rack vs $200K+ of GPU.

Sources & confidence

• Company financials: TXN 10-K FY2025 and 10-Q Q1-2026; ADI 10-K FY2025 (Oct fiscal year-end) and 10-Q Q2 FY2026 (Apr 2026); NXPI 20-F FY2025 and 20-F Q1-2026; ON 10-K FY2025 and 10-Q Q1-2026; MCHP 10-K FY2026 (Mar fiscal year-end). All filed with SEC.
• Market data: yfinance, Jun 2 2026. Prices, market caps, and derived multiples are live as of that date.
• Market size ($80–90B): est. based on WSTS global semiconductor statistics (analog segment) and IC Insights.
• EV content per vehicle ($50–80 analog): est. from TXN and NXPI investor presentations. Varies by vehicle platform.
• AI server analog content ($10–30 per rack): est. from industry teardowns.
• Confidence: Revenue, margins, FCF, debt, inventory, capex, and share counts are from SEC filings (high confidence). End-market revenue splits are approximate — companies report segment breakdowns differently. Market-size and content-per-unit figures are estimates from industry sources and company presentations, not SEC-filed data.