Will Tower Semiconductor achieve $2B or more in total revenue for fiscal year 2026?

The math: $2B requires average quarterly revenue of $500M. Q1 2026 guidance is $412M. That means Q2-Q4 must average $529M. Q4 2025 was $440M. So Q2-Q4 2026 must average 20% above Q4 2025 levels. The SiPho 5x capacity ramp is the key driver -- if tools qualify on schedule and capacity utilization reaches 75-85% at Fab 2 and Fab 9, the revenue acceleration in H2 is plausible. SiPho revenue alone could grow from the current ~$380M annualized to $600M+ if capacity ramps as planned, adding $55M+ per quarter. Combined with steady power (+20% YoY) and RFSOI recovery, $500M+ quarters in H2 are achievable. But achievement depends entirely on the capacity ramp executing on time.

Management guided QoQ revenue growth throughout 2026. If we model sequential growth: Q1 $412M -> Q2 $440M (+7%) -> Q3 $475M (+8%) -> Q4 $520M (+9%), that gives FY2026 = $1.847B -- well short of $2B. To reach $2B, the H2 acceleration needs to be steeper: Q3 $500M, Q4 $550M. This requires SiPho revenue to roughly double from Q1 to Q4 levels. Given that capacity is 5x'ing, a revenue doubling is theoretically possible but depends on tool qualification completing in Q2-Q3. Management track record of beating guidance is a positive signal, but the absolute magnitude of required growth is unprecedented for Tower.

The incremental margin analysis provides a cross-check. The 2028 model implies $2.84B revenue with incremental margins of 59% gross. If Tower is on track for this model, FY2026 should be approximately on a linear interpolation: roughly $1.566B + ($1.274B / 3) * 1 = $1.991B. That is tantalizingly close to $2B but slightly below. However, capacity ramps in semiconductors are nonlinear -- they accelerate as more tools come online. A nonlinear ramp that is slower in 2026 and faster in 2027-2028 would put FY2026 below $2B. Conversely, if the ramp is front-loaded (management seems to be pushing for this), $2B is achievable. Coin-flip probability.

I lean more cautious. The Q1 2026 guidance of $412M is a key data point -- it is BELOW Q4 2025 $440M (seasonal step-down). If Q1 is $412M, reaching $2B requires the remaining three quarters to average $529M. That is a 28% jump from Q1. While SiPho ramp could drive this, the capacity qualification market showed that all tools qualified by December 2026 has only ~38% probability. If not all tools are qualified, revenue capacity is constrained. The revenue depends directly on the capacity ramp, and the capacity ramp probability is below 50%. Therefore revenue $2B should also be below 50%, though modestly so because even partial capacity ramp contributes significant revenue.

Even without full 5x capacity, Tower can grow revenue through higher utilization at existing capacity and mix improvement. Fab 2 at 60% and Fab 9 at 65% have meaningful headroom -- going to 75-80% adds substantial SiPho wafer output without new tool qualification. Additionally, RFSOI design wins at Tier 1 module providers position for 2026 revenue contribution. Power management growing 20% YoY adds $50M. The base business contributes meaningful growth independent of the 5x target. $2B is achievable even with partial capacity ramp, but it requires everything else to go well too.

The deliberate revenue reduction in mixed signal (-18%) and discrete (-14%) creates a headwind that partially offsets growth. These segments combined lost roughly $50-60M YoY in FY2025. If this continues into 2026, it reduces the net revenue growth by ~$25-30M, which matters when the target is tight. The 2028 model assumes these reductions are complete and replaced by higher-margin flows, but the transition may not be done by 2026. Net assessment: $2B is a stretch target for 2026, more likely achieved in Q1-Q2 2027 run rate terms.

The $2B threshold requires significant H2 2026 acceleration. SiPho capacity ramp is the key swing factor. Management has guided for QoQ growth throughout 2026 and has a track record of beating guidance. Near coin-flip with slight lean toward NO given the aggressive math.

Linear interpolation to the 2028 model gives ~$1.99B for FY2026 -- essentially on the boundary. The question is whether the ramp is linear (just under $2B) or front-loaded (above $2B) or back-loaded (well below). Management emphasis on QoQ growth and 5x capacity by year-end suggests front-loading effort. But semiconductor ramps typically back-load. Lean slightly NO.

Combining the capacity qualification probability (~38%) with the base case growth, the expected FY2026 revenue lands around $1.85-1.95B in the most likely scenario. $2B requires above-expected capacity ramp and strong non-SiPho growth. Below coin-flip but not far below.