Competitive Comparison
AMD vs. Intel vs. NVIDIA
Three different bets on the same AI-accelerator demand cycle, compared on real fundamentals — a leader I don't own, a challenger I do, and a turnaround story I also hold.
| Metric | NVIDIA (NVDA) | Advanced Micro Devices (AMD) | Intel (INTC) |
|---|---|---|---|
| Market cap | ~$3.57T (source dispute; possibly higher, see model) | ~$765B (implied, 1.63B sh) | Sources conflict this quarter, mid-turnaround |
| P/E ratio (TTM) | 33.16 (forward 22.43) | 123.37 (forward 43.52) | n/m trailing (forward 76.86) |
| PEG ratio | 0.51 | 0.73 | N/A (turnaround still being priced) |
| EV/EBITDA | 29.43x | 82.00x | 23.73–43.03x (wide swing) |
| Capex (TTM) | $6.57B | $282M (quarterly) | −$13.10B |
EV/EBITDA, compared
Why I own AMD and INTC, not NVIDIA
NVIDIA is the dominant AI-GPU supplier by a wide margin, and its PEG of 0.51 is still a genuinely reasonable number for a market leader — I don't dispute that. I don't own it because the thesis is already fully priced by the entire market, not because I doubt the business. AMD is my bet that the accelerator market doesn't stay a single-vendor story forever; its own PEG has compressed sharply (0.73, down from well above 7 a cycle ago) as its earnings caught up to its multiple, though at 123x trailing earnings versus NVIDIA's 33x, the market is still charging a real premium for AMD being a credible second source rather than the leader. Intel is a different bet entirely — a manufacturing turnaround now showing its first real proof point (Q2 2026 revenue up 25%, EPS doubling estimates), but its multiples are still swinging wildly quarter to quarter as the market tries to price how durable that inflection is, which is exactly why it's sized small rather than treated as a settled thesis.