Conviction: High
Status: Holding (Core Position) — RSI ~52.9 (computed), price $180.05, -7.9% this week
Quick Snapshot
| Signal |
Reading |
| Overall |
🟡 RSI ~52.9 (computed from 19 OHLC entries) — neutral but declining from 63.3 post-earnings peak. Price $180.05, -1.3% today, -7.9% this week, +3.4% 30D. SMA19 approx ~$185.53. Price -2.9% below SMA. |
| Moat |
Wide |
| Key insight |
Post-earnings digestion continuing. Price dropped from $185.57 (02/26) to $180.05 (-2.99%) this week. The -7.9% 7D reflects the post-Q4 sell-the-news move extending. RSI declined from 63.3 to ~52.9 — healthy normalization. The 30D is still positive (+3.4%), showing the underlying trend is up. GTC Conference upcoming in March is the next major catalyst. Core thesis: $97B FCF, 75% gross margins, Blackwell ramp — all intact. |
Action Matrix
| Action |
Level |
Why |
| Current |
🔒 Hold |
Core position +178%, post-earnings digestion is normal. RSI at 53, not broken. |
| Add Zone |
$175 - $185 |
RSI < 45 pullback = add opportunity. Currently $180 is at bottom of add zone. |
| Stop-Loss |
$155 (-14%) |
Below major support, would signal structural breakdown |
| Target |
$250 (+39%) |
FY2027 revenue $300B+ at 25x forward P/E |
Price Data
| Stock |
Price |
1D |
7D |
30D |
3M |
52wkHi |
RSI |
Status |
Action |
| NVDA |
$180.05 |
-1.3% |
-7.9% |
+3.4% |
— |
-7.9% |
~53 |
🟡 Neutral/Declining |
🔒 Hold |
Note: RSI computed from 19 OHLC entries (~52.9). SMA20 field N/A; approx 19-entry average ~$185.53. 52wk high based on in-window data ($195.56).
Price vs Last Deep Dive (02/26)
| Metric |
02/26 |
03/03 |
Change |
| Price |
$185.57 |
$180.05 |
-$5.52 (-2.9%) |
| RSI |
63.3 |
~52.9 |
-10.4 (normalizing from post-earnings peak) |
| SMA20 |
$186.78 |
~$185.53 (approx) |
~flat |
| vs SMA |
-0.7% below |
~-2.9% below |
Slightly more below SMA |
| 30D |
-0.5% |
+3.4% |
Improved (base effect + slight price gain) |
Update from 02/26: Post-Q4 earnings sell-the-news digestion is continuing. Price dropped from $185.57 to $180.05 (-2.99%) while RSI normalized from 63.3 to ~52.9. The -7.9% 7D reflects the week that included the post-earnings drop and continuation. However, the 30D flipped from -0.5% to +3.4% — showing the underlying trend has turned positive over the past month. At $180.05 with RSI ~53, NVDA is in the lower part of the add zone ($175-185). GTC Conference in March should be the next catalyst. Core position maintained.
Company Overview
One-Liner
NVIDIA designs GPUs and data center-scale AI infrastructure that powers the majority of the world's AI training and inference workloads.
Business Model
| Question |
Answer |
| What they sell |
GPUs, CPUs, DPUs, networking (NVLink/InfiniBand), full-rack AI systems (Blackwell, Grace), CUDA software platform |
| Who pays |
Hyperscalers (AWS, Azure, GCP), AI model makers (OpenAI, Anthropic, Meta), enterprises, sovereign AI initiatives |
| Revenue model |
Hardware sales (chips + systems), software licensing (NVIDIA AI Enterprise, vGPU), development services |
| How sticky |
Extremely sticky — CUDA ecosystem has 7.5M+ developers, massive switching costs |
Key Segments
| Segment |
Revenue % |
Growth |
Notes |
| Data Center |
~91% |
+73% YoY (Q4) |
AI training + inference, Blackwell ramping, $62.3B in Q4 |
| Gaming |
~6% |
+30% YoY |
GeForce RTX 50 series launched |
| Professional Visualization |
~1% |
+56% YoY |
RTX PRO GPUs |
| Automotive |
~1% |
+32% YoY |
DRIVE Hyperion platform |
Competitive Analysis
Competitive Moat
| Moat Type |
Present? |
Evidence |
| Network effects |
🟢 |
7.5M+ CUDA developers, massive software ecosystem |
| Switching costs |
🟢 |
Entire AI stack (PyTorch, TensorFlow, JAX) optimized for CUDA |
| Cost advantages |
🟢 |
Best performance per watt — Blackwell delivers 4x inference throughput vs Hopper |
| Intangible assets |
🟢 |
10,000+ patents, CUDA platform (18 years of optimization) |
| Efficient scale |
🟢 |
Fabless model with TSMC's best process nodes |
Moat Width: Wide
Moat Trend: Stable (widening in systems, narrowing at chip level from custom ASIC competition)
Financials
Full Year FY2026: Revenue $215.9B (+65.5% YoY). GAAP Gross Margin 75.0% in Q4. FCF $97B for the year.
Q4 FY2026 (reported Feb 25, 2026): Record $68.1B revenue, +73% YoY. $35B FCF in a single quarter.
Key Metrics
| Metric |
Q4 FY2026 |
Q3 FY2026 |
Trend |
| Revenue |
$68.1B |
$57.0B |
📈 +19.5% QoQ |
| Gross Margin |
75.0% |
73.4% |
📈 |
| FCF |
$35B |
~$14.7B |
📈 |
| EPS (diluted) |
~$1.65 |
$1.30 |
📈 |
Bull Case
Why This Could Work
- AI infrastructure spending is accelerating — Q4 revenue $68.1B was +73% YoY; Q1 guidance crushed estimates.
- Blackwell ramping + Rubin on deck — Each generation delivers 10x+ inference improvement.
- $97B FCF, 75% gross margins — Self-funding growth machine.
- CUDA ecosystem moat deepening — 7.5M+ developers, 78% of TOP500 supercomputers.
- GTC Conference March 2026 — Expected product roadmap updates, customer wins = catalyst.
Upside Scenario
| If This Happens |
Stock Could |
| FY2027 revenue exceeds $300B |
$250-275 (+38-53%) |
| Inference demand explodes with agentic AI |
$300+ (+67%) |
Bear Case
What Could Go Wrong
- Custom ASIC competition — Google TPUs, Amazon Trainium, Meta's AMD Helios deal.
- Export controls — China data center market "effectively foreclosed" per 10-K.
- AI capex cycle slowdown — Hyperscalers pull back if AI ROI doesn't materialize.
- Heavy insider selling — -$3.0B net (0 buys, 100 sells) — routine but persistent.
Thesis Killers
- Gross margins decline below 65% for 2+ consecutive quarters
- Major hyperscaler publicly shifts >50% of AI compute to custom silicon
- Revenue growth decelerates to <20% YoY
Entry Strategy
Technical Levels
| Level |
Price |
Notes |
| Support 1 |
$175 |
Entry zone floor / strong buy territory |
| Support 2 |
$172 |
Feb low, strong buying appeared |
| Support 3 |
$155 |
Stop-loss — structural breakdown |
| Resistance 1 |
$185 |
SMA20 (approx) |
| Resistance 2 |
$195.56 |
In-window 52wk high |
| All-time high |
$207.03 |
Oct 29, 2025 |
Entry Approach
| Strategy |
Details |
| Current |
Hold core position |
| Add zone |
$175-185 with RSI < 45 |
| At $180 current |
Approaching add zone bottom |
| Stop |
$155 (-14%) |
Upcoming Catalysts
| Date |
Event |
Impact |
Watch For |
| 2026-03 (expected) |
GTC Conference |
🟢🟢 |
Product roadmap, Rubin details, customer wins |
| 2026-05 (est.) |
Q1 FY2027 Earnings |
🟢 |
Blackwell Ultra ramp, guidance trajectory |
| H2 FY2027 |
Rubin platform ships |
🟢 |
Next-gen, 10x inference improvement |
| 2026 |
USG export control replacement rule |
🟡 |
Scope and restrictions — risk factor |
Research Log
| Date |
Update |
| 2026-02-26 |
Created deep dive with full Massive financial data. Q4 FY2026 post-earnings. Price $185.57, RSI 63.3, SMA20 $186.78. Core position +178%. Hold maintained. |
| 2026-03-03 |
Post-earnings digestion continues. Price $180.05 (-2.9% from 02/26). RSI ~52.9 (computed; standard RSI field unavailable). -7.9% 7D. 30D improved to +3.4%. Price now -2.9% below SMA. At lower end of add zone. GTC Conference = next catalyst. Hold. |
The Gold
Key Discoveries
| Discovery |
Implication |
| FY2026 FCF $97B — more than most companies' total revenue |
Can self-fund any R&D, return massive capital, acquire strategically |
| Trailing P/E (39.7) is 38% below 5-year average (63.9) |
Earnings growth massively outpaced stock price |
| Post-earnings -7.9% while 30D flipped positive |
Market consolidating after beat — not broken |
| GTC Conference March 2026 = imminent catalyst |
Roadmap updates could re-catalyze the stock |
Open Questions
- What will be announced at GTC Conference March 2026?
- How does the USG export control replacement rule affect product scope?
- At what point does custom ASIC adoption reach a tipping point for inference?
- How will Rubin/Vera Rubin transition affect gross margins in FY2027?
Mistakes (If Applicable)
| Mistake |
Lesson |
| N/A — holding since ~$65, +178% |
Post-earnings digestion is normal; don't sell on 7D weakness |
Sources
- Fundamentals figures: company-reported results (quarterly/annual filings) as available at the artifact date; predates the desk's EDGAR reconciliation gate — figures not re-verified after publication.