Video summary
Nvidia's Blowout Quarter Just Made These Stocks Massive Winners
Main summary
Key takeaways
Market/sector takeaway
- Nvidia (NVDA) delivered an exceptionally strong quarter; NVDA stock rose ~6% pre-/post-earnings (after being down ~1–2% before the call).
- The semiconductor ecosystem moved higher in sympathy—“all green” across semiconductor names, with NVDA the strongest.
- The segment also includes “Neo Cloud” / AI infrastructure players; Nebius and CoreWeave were highlighted among the best performers.
Macro / capex signal (why semis rallied)
- Nvidia signaled major AI infrastructure spend:
- Capex: ~$1.3–$1.4T for next year, described as well above market expectations.
- The argument: despite being cash-intensive, this is good for semiconductors and Neo Clouds.
- Nvidia cited ROIC under one year for its “new clouds” approach—suggesting attractive economics (not every player will replicate, but it’s framed as evidence).
Nvidia earnings & guidance (key numbers)
Near-term performance
- Revenue: +106% YoY
- Gross margin: ~75%, expected to decline due to high memory prices
Forward growth
- Next quarter guidance: ~90% YoY growth
- Q3 revenue expectation: $108B (~90% YoY)
- Q3 gross margin: 74% (down from 75%)
- Management attributed the move to memory costs and the Vera Rubin ramp
Fiscal 2028 guidance (emphasis)
- Fiscal 2028 revenue growth: ~70%
- The speaker notes the market expected less, and frames upside to ~100% if demand stays unconstrained and supply allows.
- Gross margin path for fiscal 2028:
- 75% → ~74% → low ~71.5% → back to ~72.5%
Segment drivers / product roadmap (timeline & metrics)
Data center / hyperscalers
- Hyperscalers: +102% YoY
- Data center overall: +117% YoY
- “ACIE” growth: +138% YoY
- Now ~50% of the data center segment
- Expected to outgrow hyperscale growth over time
AWS / datacenter GPU expansion
- 2M additional GPUs plus Vera CPUs deployed from this quarter through Q2 FY2029 (≈ two years).
- Mentioned use cases include:
- Serving Nemotron models on Bedrock and SageMaker
- Use of Nvidia’s “physical AI stack” for Oracle / Now (as transcribed)
- Revenue potential cited:
- 2M GPUs could generate ~$200B of extra revenue (per the speaker)
Vera Rubin (production ramp)
- Production shipments began in early August.
- Orders already in from every major hyperscaler AI cloud and system OEM.
- Ramp/performance claims:
- 30x higher throughput per megawatt
- 35x lower token cost vs Grace Blackwell Ultra
- System adoption:
- Grok “LPX” rack-scale system: volume shipments to early adopters later this quarter
- Nebius named as the first Neo Cloud
- Grok performance: ~4x tokens per second of the “next best alternative” (per subtitle claim)
Vera CPU
- Revenue expected to more than double in fiscal 2028.
Networking
- Networking: +18% QoQ
- Spectrum-X Ethernet: 2.6x YoY
- Management claim (speaker paraphrase): Nvidia is the largest and fastest growing networking company.
Memory pricing & capacity risk management (disclosure of risk)
- Nvidia reset its margin guidance due to “extreme memory pricing.”
- Key framing:
- Memory scarcity is described as a symptom of the AI buildout, not an unrelated cost shock.
- Nvidia is working with all three major memory suppliers on capacity.
- The speaker counters the idea that the memory cycle is over: “think again.”
Balance sheet / commitments & financing ecosystem
Nvidia commitments (liquidity vs obligations framing)
- Total commitments: $366B
- Majority:
- $279B supply & capacity, up from $119B last quarter
- Largely tied to memory procurement
- “Frontiers lab” investment:
- Subtitle suggests ~$50 (units unclear), described as a small fraction of expected free cash flow (context indicates potentially ~$50B).
Third-party AI lab financing platforms
- Newly announced financing platforms involving:
- Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR
- Purpose:
- Mobilize $500B+ of third-party capital for AI labs
- Distinction noted:
- The $500B+ is outside lenders, not on Nvidia’s own balance sheet.
Time horizon
- Commitments expected largely over ~2.5–3 years, with visibility extending to 2032 and beyond.
Explicit “demand vs supply” framework (what limits what)
The call interpretation implies a constraint model:
- ~100% demand if “unconstrained”
- ~70% supply-constrained number
- Nvidia is portrayed as confident it can deliver above 70% if supply allows:
- 70% is framed as conservative, with potential up to ~100% in fiscal 2028 if unconstrained.
Valuation / DCF inputs & outputs (numbers)
- DCF tweak:
- Next fiscal year growth assumption:
- Bull: 75% YoY
- Base: 70% YoY
- Bear: 65% YoY
- Year 3 (fiscal 2029) growth: slightly increased
- EBIT margins reduced by ~1 percentage point:
- Bull: ~65%
- Base: ~64%
- Bear: ~60%
- Next fiscal year growth assumption:
- Valuation result:
- Probability-weighted price: $309
- Upside: 38.7% from the “price we’re at today”
Disclaimer note
- The subtitle contains an investor implication, but the provided text does not clearly include “not financial advice.”
Who benefits (ticker-level coverage)
Semiconductors / memory (direct beneficiaries)
- SK Hynix
- Micron (HBM/DRAM)
- Samsung
Foundry / test / packaging / infrastructure supply chain
- TSMC
- Amkor
- Teradyne
- Advanced packaging: Coherent (transcribed as “Coeus/Coeus”)
- Nvidia multi-year capacity deal at Amkor’s Arizona site
Networking / interconnect (direct & indirect)
- Astera Labs
- Credo
- Coherent (also appears here)
- Lumentum
- Applied Optoelectronics
- Marvell
- Arista (explicitly mentioned by Nvidia; also reports later that day)
- Amphenol
Power, cooling, and power electronics
- Vertiv
- Vizio (as transcribed)
- Evernova (as transcribed)
- Monolithic Power
- Analog Devices
- onsemi (“On Semi”)
Hyperscalers / cloud platforms & “Neo Clouds”
- AWS (Amazon Web Services)
- Microsoft
- Oracle
- Equinix
- Nebius (spelling inconsistencies: “Nebius”/“Neubias”)
- “Neo Cloud” concept tied early to CoreWeave (also referenced again)
System builders / hardware assemblers
- Dell
- Supermicro
- HPE
- Celestica
CPU/architecture mention
- Arm Holdings (noted as high-quality but premium-priced per speaker)
Other semiconductor mention
- AMD (suggested it may benefit if Nvidia can supply only some portion of demand)
Risks / cautions raised by the speaker
- Memory cost risk: gross margin pressured by memory pricing; expected to bottom ~71.5% before recovering.
- Valuation dispersion risk: some names may already price in future success, while others may still be undervalued (“priced into” caution).
- Commitment interpretation: many numbers are commitments, not immediate cash outflows; timing of ~2.5–3 years emphasized.
Methodology / framework(s) explicitly used
- Demand vs supply framework (call interpretation):
- Demand: ~100% unconstrained
- Growth guidance: ~70% supply-constrained for FY2028
- DCF scenario framework:
- Next-year growth: Bull/Base/Bear = 75% / 70% / 65%
- EBIT margin adjustments: ~65% / 64% / 60%
- Output: $309 probability-weighted and 38.7% upside
Disclosures / presenters / sources
- Presenter/author: Only the speaking individual is visible in subtitles; name not provided.
- Sources mentioned (companies/partners):
- Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, KKR
- Possible acquisition report of Hugging Face (price cited ~$13–$14)
- Nebius, CoreWeave, and hyperscalers (AWS, Microsoft, Google, Oracle)
- Memory suppliers and ecosystem companies listed above (no external research house named)