Confidence legend: ✓ verified-primary (Tower 20-F, Tower IR releases, Skyworks 10-K, IDC / Counterpoint smartphone-shipment data) · ◐ partial / aggregator · ⚠ inferred / estimate.
This page covers Tower’s largest historical end market: the handset RF front-end (RFFE) franchise. Tower’s SiGe BiCMOS, RF SOI, and 65nm RF CMOS process platforms supply power-amplifier (PA) MMICs, switches, low-noise amplifiers (LNA), and front-end-module (FEM) integration silicon to handset RFFE customers — most prominently Skyworks Solutions ✓, and at aggregator-confirmed-only levels Qorvo, Murata, and Broadcom.
The single most important framing fact: RF Mobile compressed from 29% of revenue in FY 2024 to 23% in FY 2025 (Q4 2025 slides ✓) — a $58M absolute decline year-over-year against the broader Tower revenue mix-shift toward RF Infrastructure (SiPh + SiGe driver IC). The handset RF franchise is range-bound at $360–420M annually until the envelope-tracker ramp + 5G mid-range smartphone refresh cycles deliver material recovery in 2026–2027.
1. Smartphone Cycle Context
The global smartphone unit-shipment cycle (per IDC Quarterly Smartphone Tracker and Counterpoint Research ◐):
| Year | Global smartphone units (M) | YoY growth | Note |
|---|---|---|---|
| 2021 | ~1,355 | +6% | Post-COVID rebound |
| 2022 | ~1,205 | -11% | China zero-COVID + global inflation |
| 2023 | ~1,167 | -3% | Trough |
| 2024 | ~1,240 | +6% | Modest rebound; AI-phone narrative |
| 2025 | ~1,235 | -0.4% | Plateau; AI-phone marketing not driving upgrade volumes |
| 2026 | ⚠ ~1,260-1,290 | +2-4% | Range-bound consensus |
⚠ Specific 2025 / 2026 figures vary by tracker; the structural pattern is the same: the global smartphone market is mature and range-bound at ~1.2B annual units, with growth concentrated in emerging markets (India, Southeast Asia, Africa) at low ASPs and shrinking in mature markets (North America, Europe, China, Japan, Korea) at high ASPs.
For Tower’s RF Mobile revenue, the smartphone unit cycle is the primary upstream driver, but with two important offsets:
- 5G content-per-phone is rising even as unit volume plateaus. A 5G premium-tier handset contains 8–12 PA / FEM dies vs ~4–6 in a 4G-only handset. The mid/high-band 5G architecture (n77, n78, n79, mmWave bands) requires SiGe BiCMOS or RF SOI process complexity that Tower supplies.
- Tower revenue lags handset shipments by 6–9 months as wafer flow precedes module assembly which precedes handset assembly which precedes channel-fill which precedes consumer purchase. A Q4 2024 smartphone shipment cycle drove Q1–Q2 2024 Tower wafer revenue. The 2024 IDC + Counterpoint shipment data is therefore the read-through to Tower’s 2024 RF Mobile revenue trajectory.
✓ Skyworks Solutions explicitly attributes its Q1 2026 results commentary to handset cycle stabilization and the iPhone 17 ramp; analogous sell-side commentary applies to Tower’s RF Mobile revenue with the lag.
2. Tower RF Mobile Customer Dynamics
Skyworks Solutions — confirmed primary customer (✓)
Skyworks is the single most clearly-confirmed named Tower customer in the RF segment:
- 2012: Tower (then TowerJazz) named Skyworks Solutions 2012 Foundry Supplier of the Year + Skyworks “Quality Iron Man Partner Award” (TowerJazz IR release ✓).
- Multi-decade relationship — Skyworks uses Tower’s SiGe BiCMOS process families (TS18SL, SBC18S5) for power-amplifier-based products in volume production.
- Skyworks’s primary end customer is Apple (iPhone). Skyworks’s other customers include premium Android OEMs (Samsung, Google, OPPO, Vivo, Xiaomi).
- Skyworks’s foundry supply chain includes Tower (for SiGe BiCMOS), TSMC (for RF SOI on advanced nodes), and Skyworks’s own gallium-arsenide (GaAs) backend in Newbury Park CA.
⚠ The 2025–2026 announced Skyworks-Qorvo merger (Knowmade — RF technologies merger ◐) creates uncertainty about post-merger foundry-allocation patterns. Whether the merged entity retains Tower’s full Skyworks-legacy wafer flow, consolidates onto fewer foundry partners, or shifts share to Qorvo’s pre-merger foundry preferences is not yet public. Watch-item for the next Tower 6-K and the post-merger entity’s first investor disclosures.
Qorvo — aggregator-cited (⚠)
Qorvo (formed 2015 from RFMD + TriQuint merger) is not named in any Tower primary-source disclosure I could locate. Industry-knowledge-level expectation is that Qorvo sources some SiGe / BiCMOS volume from Tower among multiple foundry partners — but this is NOT primary-source-confirmed.
- Qorvo’s own 10-K Item 1 / Risk Factors reference foundry-partner concentration risk but do not name specific foundries.
- Tower’s historical PRs include Qorvo design-win-class announcements at the trade-press level only.
⚠ TODO: cross-check Qorvo’s most recent 10-K filings for any foundry-partner naming; cross-check Tower historical PRs for Qorvo-specific design-win announcements.
Murata Manufacturing — aggregator-cited (⚠)
Murata’s RF-FEM business (post-Peregrine acquisition / SAW + RF-IC integration) is not named in Tower primary-source disclosure. Industry-knowledge-level expectation is some specialty SiGe / RF wafer flow at Tower among multiple foundry partners — but this is NOT primary-source-confirmed.
⚠ TODO: cross-check Murata investor disclosures for Tower-specific naming.
Broadcom — aggregator-cited (⚠)
Broadcom’s RFFE business (post-Avago / Broadcom integration) is generally structured as IDM with internal fab capacity for the GaAs / FBAR portions and external foundry sourcing for the SiGe / SOI portions. Specific Tower-Broadcom RFFE wafer flows are aggregator-cited only.
Other RF customers cited in aggregator sources
Wikipedia and industry-aggregator articles historically list Semtech, Vishay, Teledyne as Tower customers in the analog/RF/specialty segment. Without a Tower-disclosed or customer-disclosed primary-source cross-confirmation, treat all such names as ⚠ aggregator-level.
A directly-verified new RF design win (defense / SATCOM segment, not handset RFFE):
- Axiro Semiconductor — high-power, high-efficiency SiGe ICs for U.S. defense applications; co-disclosed by Tower 2026-04-27 (Tower-Axiro PR ✓).
3. The Apple iPhone Envelope-Tracker Design Win (the FY 2026–2027 Anchor)
The Q4 2025 Tower earnings call cited a “previously announced tier-one handset envelope tracker” 300mm ramp expected to gain share over time (Investing.com Q4 2025 slides ✓). The envelope tracker is widely understood to be an Apple iPhone PMIC component based on industry triangulation:
- The “tier-one handset” framing in Tower disclosures has historically referenced Apple-related design wins (Apple is the single largest premium-handset OEM globally).
- The Tower-ST Agrate 300mm BCD process is the production node for premium PMIC + envelope-tracker ASICs.
- Apple has historically used external foundry sourcing for non-A-series handset silicon (the A-series APs are TSMC; PMIC is mixed; envelope tracker is mixed).
⚠ Apple is NOT customer-named in Tower disclosures; the customer-name attribution is aggregator-level / industry-triangulation-level inference.
The envelope-tracker design-win significance:
- Single-sourced design win at a premium-volume tier-one customer creates multi-year revenue visibility (typical iPhone PMIC design lifespan: 2–3 generations + extended-tail manufacturing for legacy SKUs).
- 300mm Agrate capacity is the production node — meaningfully higher per-wafer ASP than 200mm equivalents.
- “Gain share over time” Tower commentary implies the iPhone PMIC sockets ramp from initial-design-win volume in 2026 to broader allocation in 2027–2028.
The envelope-tracker design win is the single most important specific watch-item for the RF Mobile recovery thesis — it anchors a $50–150M incremental revenue trajectory through 2026–2028 (⚠ analyst estimate).
4. 5G-vs-4G Content Per Phone
Premium-tier 5G handsets contain meaningfully more PA / FEM die than 4G-only handsets:
| Generation | Bands | PA dies / FEM | Cumulative wafer area |
|---|---|---|---|
| 4G LTE | 6–10 bands | 4–6 PAs | Baseline |
| 5G sub-6 GHz | 8–14 bands | 8–10 PAs | ~1.5× 4G baseline |
| 5G sub-6 + mmWave | 10–18 bands | 8–12 PAs + mmWave FEMs | ~2× 4G baseline |
| 5G CA + 4G fallback | All bands | 10–14 PAs | ~2.5× 4G baseline |
⚠ Specific PA-count varies by handset OEM and tier; the structural trend is that the 5G-mmWave premium tier roughly doubles the per-handset RF wafer content vs 4G LTE.
For Tower’s wafer-volume trajectory, the 5G-content-per-phone uplift is a multi-year tailwind even as unit shipments stagnate. The math: even if global handset units are flat at 1.2B annually, the average wafer content per handset is rising at ~5–8% CAGR through 2026–2028 as 5G penetrates further into mid-tier and emerging-market handsets.
PA architecture trends
The handset PA architecture is migrating across multiple dimensions:
- Envelope Tracking (ET) — dynamically adjusts PA supply voltage to match the modulated signal envelope. Reduces PA power consumption by 20–40% vs fixed-supply operation. Requires precision PMIC + envelope-tracker IC. Tower’s BCD process supplies envelope-tracker ASICs (the tier-one design win).
- Digital Predistortion (DPD) — corrects PA nonlinearity in the digital domain. Implemented in baseband + RF mixed-signal silicon (typically not Tower’s scope, but indirect demand signal).
- Integrated SiGe BiCMOS modules — combines PA + driver + LNA + switch on a single SiGe BiCMOS die. Tower’s SiGe BiCMOS process is a primary foundry option for these modules. The mid-tier handset shift toward integrated SiGe modules vs discrete-die approaches is structurally favorable for Tower wafer demand.
- GaN-on-SiC at high-band — for mmWave and emerging 6G bands, GaN-on-SiC may displace SiGe at the high-end. Tower has historically emphasized SiGe rather than GaN; this is a structural watch-item for 6G transition timing.
The mid-tier shift toward integrated SiGe BiCMOS modules vs discrete plays is the single most important architectural trend for Tower’s RF Mobile revenue. If integrated modules grow faster than discrete-die solutions, Tower’s per-handset wafer content rises faster than the headline 5G-content-per-phone trend implies.
5. RF Mobile Cycle Status (2022–2026)
2022: Peak utilization
Post-COVID handset cycle peak; Tower RF Mobile utilization >85% across SiGe + RF SOI platforms; ASPs supported by industry-wide chip shortage.
2023: Inventory correction
Sharp Q1–Q2 2023 inventory destocking across Skyworks, Qorvo, Broadcom, Murata; Tower wafer-starts dropped meaningfully; FY 2023 RF Mobile revenue contracted vs FY 2022. 5G premium-tier shipments held up; mid-tier and Chinese-OEM volumes weakened.
2024: Stabilization
Inventory-rebuild signal from RFFE customers; Tower wafer flow recovered through H2 2024; FY 2024 RF Mobile = ~29% of $1.44B = ~$418M (segment revenue mix table).
2025: Compression vs broader mix
FY 2025 RF Mobile compressed to 23% of $1.57B = ~$360M (-14% YoY). The compression reflects:
- Mature smartphone unit-volume cycle (-0.4% global units in 2025).
- Chinese OEM ASP pressure on Tower’s foundry pricing.
- Mix-shift dynamics — Tower RF Infrastructure surged while RF Mobile held flat-to-down.
Q4 2025 mix: RF Mobile = 24% of Q4 2025 revenue (slight QoQ stabilization vs full-year 23% mix).
2026: Forward outlook
Tower management framing on Q4 2025 call: “sequential revenue and profitability growth throughout 2026” with the envelope-tracker ramp + Chinese-OEM 5G mid-tier refresh providing the recovery anchors.
⚠ Analyst-class FY 2026 RF Mobile estimate: ~$400–450M, recovering toward FY 2024 baseline as the envelope-tracker ramp adds incremental revenue and the underlying handset cycle stabilizes.
2027–2028: Recovery scenario
The Apple envelope-tracker design win + 6G research-tier silicon + continued integrated-SiGe-module penetration imply RF Mobile can recover to ~$500–600M by FY 2028 (~18–21% of $2.84B FY 2028 model).
6. 6G Outlook (2030+)
6G commercial deployment is 2030+; current 6G-related wafer volume is research-tier only. The 6G-related RF wafer demand for Tower:
- High-band SiGe — 6G operates above 100 GHz (sub-THz / D-band / W-band). Tower’s SBC18S5 (250 GHz fT SiGe BiCMOS) is positioned for this band; commercial volumes are 2028+ at earliest.
- 65nm RF CMOS — for 6G mixed-signal baseband + low-band RF.
- Above-100 GHz mmWave PAs — likely GaN-on-SiC at the high-power end; Tower’s GaN positioning is less developed than competitors.
⚠ 6G content forecast for Tower: minimal through 2028; ramping 2029–2032; meaningful FY 2030+ revenue contribution. 6G is not a near-term thesis driver for Tower; it is a multi-decade optionality lever.
7. Reading the RF Mobile Cycle Signals for Tower
For an analyst tracking Tower’s RF Mobile franchise, the highest-value signals to monitor:
- IDC + Counterpoint quarterly smartphone shipment data — leading indicator with 6–9-month lag to Tower wafer revenue.
- Skyworks Solutions quarterly earnings commentary — direct primary-customer signal.
- Qorvo + Murata + Broadcom RFFE quarterly earnings commentary — secondary-customer signals (aggregator-level Tower attribution but reliable cycle indicators).
- Apple iPhone shipment cadence — indirect via Skyworks; the envelope-tracker design win is iPhone-related per industry triangulation (⚠).
- Skyworks-Qorvo merger integration progress — post-merger foundry-allocation patterns will reshape Tower wafer flow.
- Tower quarterly RF Mobile revenue print — currently at $90–100M/quarter run-rate; first envelope-tracker-ramp inflection visible Q2–Q3 2026.
- 5G mid-tier penetration in China + India — emerging-market 5G is the unit-volume-growth driver as mature markets plateau.
- Above-100 GHz / 6G research-program announcements — leading indicator for the 2028+ 6G commercial trajectory.
The single most important leading indicator: Tower’s Q1 2026 / Q2 2026 RF Mobile revenue print + envelope-tracker ramp commentary. The recovery thesis requires RF Mobile to inflect higher in 2026; any further compression below the FY 2025 $360M base would compress the FY 2028 $2.84B target.
Cross-section pointers
- foundry industry dynamics — Specialty-foundry market structure context.
- AI capex cycle — Why RF Mobile is NOT primarily AI-driven (AI-cycle pull-through goes to RF Infrastructure).
- silicon photonics market — The RF Infrastructure surge is the offset to RF Mobile compression.
- TAM / SAM — Bottom-up segment-by-segment sizing.
- PH18 process — SiPh process detail.
- customers — diversified — Skyworks confirmed; Qorvo / Murata / Broadcom / Semtech / Vishay / Teledyne aggregator-level.
- Tower-ST Agrate 300mm — 300mm BCD process for the envelope-tracker design win.
- segment revenue mix — FY 2024 → FY 2025 RF Mobile compression detail.