← Back to NVDA Q1 FY27 pre-print brief
⏳ Scoring Opens May 27 — Q1 FY27 Print Window AMC
Brief published: Apr 29 · Print: May 27 · Updated same night
NVDA Post-Print Recap Semiconductors / AI Infrastructure Brief published: Apr 29, 2026 · Print: May 27, 2026

NVIDIA Q1 FY27 — The Scorecard

We published our pre-print thesis on April 29. Here's the full setup — the 5 angles we flagged the Street was missing, the comparison table, and what we'll update within 24 hours of the May 27 close.

Overall Verdict ⏳ Pending May 27
Revenue Call
⏳
Called $80–81B · Consensus $78.4B
Q2 Guide Call
⏳
Called $85–87B · Consensus ~$86B
Angles Scored
0 / 5
All pending May 27 print
01 / The Scorecard

The Call vs. The Print

What our brief said vs. what NVIDIA will report for Q1 FY27. Actual column populates after May 27 AMC. Our pre-print call was $80–81B on revenue vs. consensus $78.4B — a 2–3% above-consensus position built on contracted sovereign AI backlog and Blackwell GB300 yield improvements not in consensus models.

Metric Our Pre-Print Call Street Consensus Actual Q1 FY27 Verdict
Total Revenue
Q1 FY27 (Feb–Apr 2026)
 $80–81B
Beat-and-raise setup		 $78.4B
Mgmt guided $78B ±2%		 Pending May 27
Updates same night		 ⏳ Pending
Data Center Revenue
GB300 Blackwell systems + networking
 ~$73B
Range: $72–75B		 ~$71B
Excludes China DC		 Pending May 27 ⏳ Pending
Non-GAAP Gross Margin
GB300 yield-adjusted
 75.0–75.5%
GB300 yield tailwind		 75.0%
Guided 75.0% ±50bps		 Pending May 27 ⏳ Pending
Non-GAAP EPS
Q1 FY27
 ~$1.80–1.85
Revenue upside flow-through		 $1.76
Range: $1.74–$1.78		 Pending May 27 ⏳ Pending
Q2 FY27 Revenue Guide
The number that moves the stock
 $85–87B
Sovereign AI delivery + Q2 ramp		 ~$86B
Buy-side needs $86B+		 Pending May 27 ⏳ Pending
China DC Revenue
The free option — $0 in guidance
 $0–3B upside
H20 partial restoration possibility		 $0
Explicitly excluded from guide		 Pending May 27 ⏳ Pending
⏳ Actuals populate after NVIDIA reports Q1 FY27 on May 27, 2026 AMC. All estimates from Vektor pre-print brief (Apr 29, 2026). Consensus from Wells Fargo NVDA pre-print note (Apr 25, 2026) and Bloomberg consensus.
02 / Differentiated Thesis

5 Street Missing Angles — Scoring Pending

Our brief identified five vectors where buy-side models were absent or systematically underweighted. Each scores as ✅ Correct / ⚠️ Mixed / ❌ Missed based on what Jensen says on the May 27 call. Evidence appended same night.

Angle 01
China upside is a free option
⏳ Pending
Our pre-print thesis: The $78B guidance baked in exactly $0 in China Data Center revenue. The H20 was partially restored under a 15% US government fee structure. China historically represented 12–20% of NVIDIA revenue. Every $1B of China DC flowing through adds ~$0.04 to non-GAAP EPS — and the market wasn't pricing this optionality.
// What we're scoring against May 27
✅ Correct if: Any China DC compute revenue recognized in Q1 actuals, OR Jensen explicitly references China H20 shipments in Q2 guide.
⚠️ Mixed if: No China revenue in Q1 but positive commentary on future approvals.
❌ Missed if: China commentary is negative (new restrictions, no path forward).
Angle 02
Networking is structurally underappreciated
⏳ Pending
Our pre-print thesis: Q4 FY26 networking hit $10.98B (+263% YoY). Spectrum-X Ethernet reached parity with InfiniBand for new AI cluster deployments. Sell-side embeds networking as a Data Center line item without modeling the growth rate or attach dynamics. As networking share within Data Center expands, it's margin-accretive and undermodeled.
// What we're scoring against May 27
✅ Correct if: Networking as % of DC revenue exceeds 18% in Q1, OR Jensen specifically calls out Spectrum-X attach rate expansion.
⚠️ Mixed if: Networking grows but flattens as % of DC (mix shift toward compute).
❌ Missed if: Networking revenue declines QoQ or Jensen signals InfiniBand share loss.
Angle 03
Sovereign demand is contracted backlog, not pipeline
⏳ Pending
Our pre-print thesis: Street models treated sovereign AI as a vague TAM. It's contracted backlog. HUMAIN (Saudi Arabia) committed to 600,000 GPUs. Stargate UAE (G42) has a 1GW facility underway. Aggregate committed backlog estimated at $500B through end CY26. Revenue recognition flows Q1–Q4 FY27 as GB300 systems are racked and delivered.
// What we're scoring against May 27
✅ Correct if: Jensen quotes specific sovereign AI revenue contribution in Q1 (e.g., "sovereign contributed $X billion"), OR explicitly names HUMAIN/UAE delivery milestones.
⚠️ Mixed if: Sovereign mentioned positively but no revenue specificity.
❌ Missed if: Sovereign AI delivery slips or contracts are discussed as uncertain pipeline.
Angle 04
Blackwell GB300 cost structure improving faster than modeled
⏳ Pending
Our pre-print thesis: GB300 (Blackwell Ultra) represented ~two-thirds of Blackwell revenue in Q4 FY26. TSMC CoWoS-L packaging yields on the NVL72 configuration are maturing — early production had meaningful scrap rates. As yield improves, unit cost declines while ASP holds. The 75.0% GM guide is conservative by 50–100 bps on a yield-adjusted basis if Q1 production ran 45,000–55,000 NVL72 racks.
// What we're scoring against May 27
✅ Correct if: Reported non-GAAP GM above 75.2%, OR Kress (CFO) references GB300 yield improvement as a tailwind.
⚠️ Mixed if: GM in line at 74.8–75.2% with neutral yield commentary.
❌ Missed if: GM below 74.5% due to yield issues or mix headwinds from sovereign system deliveries.
Angle 05
Gross margin trajectory — inference scaling as independent demand curve
⏳ Pending
Our pre-print thesis: The market modeled AI chip demand primarily as training. Reasoning models (long-chain-of-thought) create a second independent demand curve through inference. A single complex query consumes 10–100x the compute of a standard inference call. Hyperscaler utilization near 100% — not supply constrained, but inference demand expanding to fill every GPU. This expands the sustainable demand ceiling and supports long-run margin stability above 74%.
// What we're scoring against May 27
✅ Correct if: Jensen specifically discusses inference demand growth as a distinct driver (not training), OR Q2 GM guide is ≥75.0% implying margin sustainability.
⚠️ Mixed if: Inference mentioned but framed as equal to (not above) training demand.
❌ Missed if: GM guide for Q2 below 74.5%, OR Jensen flags mix-shift toward lower-margin inference workloads.
Scoring methodology: Each angle scores ✅ Correct if the earnings call or reported financials confirm the thesis with specific evidence. ⚠️ Mixed if directionally right but with material qualifications. ❌ Missed if the evidence contradicts the thesis or the call goes silent on the factor entirely. Scores update within 24 hours of May 27 AMC.
03 / Scenario Setup

Pre-Print Bull / Bear / Base

Three scenarios from our brief entering the May 27 print. The Q2 FY27 guide is the primary stock catalyst — more than Q1 actuals. Buy-side needs $86B+ for Q2 to re-rate the stock. Scoring pending after print.

🟢 Bull Case
35% probability
Rev: $82–84B · DC: $74–76B · GM: 75.5% · Q2 Guide: $88–92B
→ China DC revenue flows through ($2–4B upside vs $0 guide)
→ Sovereign AI contributes specific Q1 dollar amounts
→ Q2 guide $88B+ with GM guide ≥75.5% (yield improvement)
→ Expected stock move: +10–15% AH
🔵 Base Case
45% probability
Rev: $78–81B · DC: $70–73B · GM: 74.8–75.2% · Q2 Guide: $83–87B
→ Beats $78.4B consensus modestly on Blackwell volume
→ No China revenue in Q1; positive language on outlook
→ Q2 guide in-line with Street ($83–86B) — muted reaction
→ Expected stock move: +2–7% AH (below expectations bar)
🔴 Bear Case
20% probability
Rev: <$77B · DC: <$68B · GM: <74% · Q2 Guide: $78–82B
→ GM below 74% — signals GB300 yield degradation or system mix
→ Negative China commentary (new export controls, no H20 path)
→ Q2 guide misses Street badly (<$83B) — structural demand question
→ Expected stock move: -8–15% AH
The single number that matters most: The Q2 FY27 revenue guide is the primary stock catalyst — more than Q1 actuals. A Q2 guide at $85B+ signals acceleration and re-rates the stock. A Q2 guide at $83–84B (in-line) is neutral to slightly negative given 44 of 49 analysts are at Strong Buy entering the print. Gross margin below 74.0% — signaling GB300 yield or mix issues — is the bear case trigger independent of revenue.
04 / Post-Print Assessment

What We Got Right / What We Missed / What Changed Our Model

This section populates after the May 27 print. We'll be honest about misses — that's what builds trust. A scorecard that only publishes the hits isn't a scorecard.

✅ What We Got Right
Updating May 27 AMC. We'll score each call — revenue, data center, gross margin, Q2 guide — and note which were right with precision. Every angle that fires gets documented with the exact quote from Jensen or Kress that confirms it.
❌ What We Missed
Updating May 27 AMC. If the print contradicts our thesis — directionally or on magnitude — we'll state it plainly. No burying the miss in a paragraph of hedges. The miss section is the most important part of an honest scorecard.
🔄 What Changed Our Model
Updating May 27 AMC. New information that materially shifts our Q2 FY27 and FY27 annual model — sovereign delivery pace, China policy trajectory, GB300 yield data from CFO commentary, inference utilization signals from hyperscaler call-outs.
05 / Our Pre-Print Call

Our Call Entering May 27

The specific, falsifiable pre-print call from our brief. We'll hold ourselves to exactly this after the print — not a generalized "we were bullish" claim.

// BEAT AND RAISE — pre-print verdict

Q1 FY27 Revenue: $80–81B · Q2 FY27 Guide: $85–87B · Overall: Beat-and-Raise

The $78B guide is a floor, not a ceiling. NVIDIA baked in zero China revenue — any China commentary is pure upside. The Blackwell GB300 production ramp, sovereign AI contracted backlog (HUMAIN 600K GPUs, UAE Stargate 1GW), and inference demand expanding independent of training all argue for Q1 to come in above the consensus range. The real question is whether Q2 guidance clears $85B, which is what the buy-side needs to re-rate the stock from its current level. Our base case: Q1 beats at $80–81B, Q2 guided $85–87B, stock +5–10% AH on a clean beat-and-raise without the multiple contraction scenario.

Pre-Print Brief
Published April 29, 2026 at stratify-6.polsia.app/briefs/nvidia
Report Updates
This page updates within 24 hours of the May 27 AMC print with full scoring.
NVDA Recap · Notify Me

Get the NVDA scorecard
the night it prints

This page updates May 27 within hours of the close. Drop your email to get the full NVDA Q1 FY27 recap — 5 angles scored, pre-print call vs. actual, and the forward thesis for Q2 FY27.

No spam. Unsubscribe anytime. One email when the NVDA recap drops.
← NVDA pre-print brief · Full track record · Browse all briefs