Management pitched a $90 billion 2027 roadmap but refused to answer a single question about 2026.
Thesis: This is a 'Capacity Ceiling' story disguised as infinite growth. The entire $60-90B bull case rests on three hyperscalers quadrupling spend while TSMC CoWoS capacity remains the hard bottleneck. Management is selling a 2027 dream to distract from 2025 margin dilution in semiconductors. When a CEO insults analysts for asking about margins, check your wallet.
Verdict: AVOID — Conviction: HIGH
Catalyst: The 3nm XPU ramp in H2 2025. This is the execution test for the backlog claims.
Key Risk: Customer concentration. If one of the three hyperscalers shifts architecture or cuts capex, the 2027 SAM evaporates.
The Tell: The timeline gap. Tan aggressively promoted a 2027 SAM of $60-90B but when asked for 2026 visibility to bridge the gap, he snapped: 'I don't have... good enough line of sight to want to share it with you.' You don't have visibility for next year, but you're certain about three years out? That's sales, not guidance.
Friction Level: HIGH_FRICTION — The $60-90B FY27 SAM. Bulls see a locked roadmap; Bears see three customers who need to 4x their spend in a declining capex growth environment.
Report not found
The report data is no longer available. Please return to the archive.