Finding 1 · Concentration Risk HIGH PRIORITY
Automotive end-market is 20.8 percentage points over its strategic target share — the single largest revenue-mix risk.
Evidence: FY25 Automotive revenue $1.81B (48.8% of $3.71B total) vs 28% MARKET_SHARE_TARGET from DIM_END_MARKET. Gap has widened FY23→FY25 as Auto grew ~27% while total grew ~24%.
So what: A single-market auto slowdown (regulatory, EV demand pause, tariff) would translate to a 15-20% total revenue shock. Diversification into COMM (target 15%, actual 7.3%) and Industrial (22% target, 12.2% actual) would materially de-risk the book.
Finding 2 · Customer Concentration HIGH PRIORITY
One customer — Stellar Dynamics — contributes 22% of FY25 shipment revenue ($814M).
Evidence: Top 5 customers = 45.2% of FY25 revenue. Top 20 = 65.3%. Stellar Dynamics alone (Tier 1 OEM, South region) is 4× the size of #2 account (Onyx Devices $365M).
So what: Stellar defection or major budget cut would cost $814M of revenue — nearly 2 years of organic growth. Recommend a dedicated retention plan plus accelerated growth in Tier 2 OEM and Design House segments (combined only 20% of revenue today).
Finding 3 · Product Pareto
MCU category dominates at 82% of FY25 revenue; single SKU (MCU-8-B0001) = 19.4%.
Evidence: MCU $3.06B of $3.71B. Top 15 products (5% of 284 SKUs) = 66% of revenue. Highest-margin categories (PWR 55%, RF 55%, FPGA 55%) represent only 11% of total revenue.
So what: Revenue is heavily tied to mature MCU SKUs with mid-tier margin (~50%). Growth margin-accretive categories (PWR/RF/FPGA) would materially lift blended GM% if scaled. Current trajectory keeps GM% pinned at ~50.3%.
Finding 4 · Cycle Recovery Confirmed GOOD
B:B ratio recovered from 0.94 (FY23 correction) to 1.08 (FY25), with forward momentum.
Evidence: FACT_BACKLOG authoritative B:B: FY21 1.20 (shortage peak), FY22 1.18, FY23 0.94 (correction), FY24 0.98 (early recovery), FY25 1.08 (confirmed), FY26 partial 1.13 (accelerating).
So what: Demand cycle has turned decisively. The FY23 inventory correction is fully absorbed. Continue capacity ramp plans — FY26 bookings are tracking ahead of FY25 shipments, implying shipment growth through FY26-27.
Finding 5 · External Fab Dependency
84% of shipped revenue comes from external foundries (TSMC, Samsung, GF, UMC, etc.).
Evidence: FY25 revenue by fab type: External Foundry $3.13B (84.4%), Internal $531M (14.3%), Specialty $27M (0.7%), Assembly/Test $22M (0.6%). Leading-edge (5-7nm) exclusively external.
So what: Strategic exposure to foundry allocation, geopolitics, and CAPEX cycles beyond our control. Internal fab capex should target Advanced/Mainstream nodes where internal cost-per-die is competitive. Specialty (SiC) is a 2nd-sourcing opportunity.
Finding 6 · Pipeline Health GOOD
Weighted design-win pipeline = $1.24B · late-stage conversion ($468M at Qual+Prod) covers 12% of annual revenue.
Evidence: FACT_DESIGN_WINS 10,442 records across 7 stages. Weighted by win-probability: Prospect $48M, Engage $117M, Eval $236M, Design $282M, Qual $257M, Prod $212M. Competitive displacement 64.7% vs greenfield 35.3%. Top displaced: STMicro ($128M weighted), Qualcomm ($116M).
So what: Pipeline signals solid 2-3 year forward revenue, heavily displacement-driven (good: we take share; flag: dependent on competitor weakness). Auto pipeline at $325M weighted continues to over-index vs target mix — re-examine end-market allocation of R&D.
Finding 7 · Fab Cost Efficiency
Best $/die efficiency sits in the 22–40nm band — 100% external today. Internal fabs stranded at 74% utilization on Legacy/Mature nodes.
Evidence: $/die by node (FY25): 5nm $18.40 · 7nm $14.80 · 14nm $9.20 · 22nm $6.40 · 40nm $4.10 · 65nm $2.80 · 90nm $2.10 · 130nm $1.60 · 180nm $1.20. Yield 88-96% across the stack. Internal fab util: Chandler 78% · Gresham 73% · Colorado Springs 71%.
So what: Capital is tied up on mature nodes that produce our lowest-margin products, while the efficient 22-40nm band is 100% outsourced. A node-migration feasibility study for internal fabs — even to 40nm — is the single highest-leverage capital decision.
Finding 8 · Design-Win Conversion Velocity
Average design-in → first revenue cycle is 14.2 months. Biggest leak: 41% loss at Eval → Design.
Evidence: Stage dwell (months, FY23-25 cohort): Prospect 2.1 · Engage 2.8 · Eval 3.1 · Design 4.3 · Qual 4.8. Conversion rates: Prospect→Engage 82% · Engage→Eval 74% · Eval→Design 59% · Design→Qual 71% · Qual→Prod 78%. RF/FPGA cycle 17-18mo (vs MCU 12.4mo).
So what: Eval → Design is the biggest pipeline leak at 41% loss — likely under-matched FAE coverage. Lifting Eval conversion to 70% could unlock ~$300M of incremental FY27 revenue from the current pipeline.
AutoExplore Evidence Dashboard · xFalcon AnalyticsPro · Generated 2026-04-21 · Extended 2026-04-23 (Findings 7 & 8)