Key Findings
The model argues operating margin is constrained by a high R&D run-rate (Q2 2026 R&D at $691M versus Q3 2025's $533M) and elevated stock-based compensation ($265M in Q2 2026, modeled higher sequentially). Non-operating income volatility adds further uncertainty.
Thesis
GPT acknowledges the core royalty engine benefits from infrastructure/AI mix but argues the quarter is unlikely to feature an outsized licensing recognition event. Rising deferred revenue supports steadier licensing visibility but 'does not guarantee a discrete, EPS-boosting quarter.'
What This Means
This creates a stark divergence with Gemini at $0.51 (+24% vs consensus). The AI models fundamentally disagree on whether ARM's infrastructure royalty acceleration can overcome cost pressures. Investors face a wide range of outcomes depending on licensing timing and expense management.