MRVL vs NVDA: AI Stock Comparison
Marvell Technology, Inc. vs NVIDIA Corporation: AI exposure, key metrics, next earnings dates, upcoming catalysts and risks — every number linked to its source.
$NVDA scores higher on AI exposure (5/5 vs 4/5) under our published rubric. The score measures how directly each business is tied to AI, not which stock to buy.
⏱ Next catalyst: $NVDA has the earlier next catalyst (Nov 15, 2026).
| AI exposure score | ⚡⚡⚡⚡⚡ 4/5 | ⚡⚡⚡⚡⚡ 5/5 |
|---|---|---|
| Shares outstanding | 876.9M as of 2026-08-21 · 10-Q cover page · SEC filing | 24.10B as of 2026-08-21 · 10-Q cover page · SEC filing |
| Category | AI chips | AI chips |
| Key metrics | Revenue (latest quarter) $2.74B 10-Q period ended 2026-08-01; +37% y/y · SEC filing Gross margin 53.1% Gross profit / revenue, same quarter · SEC filing Operating income $459.7M Same quarter; 17% operating margin · SEC filing | Revenue (latest quarter) $96.22B 10-Q period ended 2026-07-26; +106% y/y · SEC filing Gross margin 75.0% Gross profit / revenue, same quarter · SEC filing Operating income $63.73B Same quarter; 66% operating margin · SEC filing |
| Next earnings | ||
| Next catalyst | Nov 15, 2026 – Nov 20, 2026 SC26 Supercomputing Conference · source | |
| Upcoming catalysts | Nov 15, 2026 – Nov 20, 2026 SC26 Supercomputing Conference · source Dec 8, 2026 – Dec 13, 2026 NeurIPS 2026 (Atlanta; also Sydney & Paris) · source | |
| Most recent reported | Nov 2, 2026 Marvell to acquire Lumentum in all-stock deal (~$115B EV) · source Combines electrical and optical AI connectivity; expected to close 1H 2028 pending approvals. | Nov 4, 2026 CUDA 13.5 ships with Vera Rubin general-availability support · source Software readiness milestone for NVIDIA's next GPU architecture ahead of volume deployments. |
| Key risks |
|
|
| AI angle | Marvell's data-center segment (custom AI accelerators, electro-optics, switching) is its growth engine. The planned Lumentum acquisition, announced Nov 2, 2026, would add lasers and transceivers. | NVIDIA's data-center GPUs, NVLink/InfiniBand networking and CUDA software are the default stack for training and inference. Hyperscaler and neocloud capex plans are the main swing factor for its results. |
| Full analysis: $MRVL → | Full analysis: $NVDA → |
The morning brief
Today's AI-stock catalysts in your inbox before the open. Free, 2-minute read.