Nvidia’s 2027 Earnings Calls: ACINE Growth and Supply Constraints Contradict

Date of Call: Aug 26, 2026

Financials Results

• Revenue: $96B, more than doubled year over year
• EPS: Record EPS, not explicitly quantified
• Gross Margin: 75%, largely unchanged from last quarter

Guidance:

• Q3 revenue expected to be $108B +/- 2%, driven by ACINE with data center growth to re-accelerate in Q4.
• Fiscal Year 28 revenue expected to grow approximately 70% year over year.
• Q3 gross margin expected to be 74% +/- 50 bps; expected to bottom in Q4 (71-72%) and settle at 72-73% in FY28.
• Full-year FY27 operating expenses expected to grow in the low 50s.
• Tax rates expected between 16-18% for FY27.

Business Commentary:

Record Revenue Growth:

• NVIDIA reported record revenue of $96 billion for the second quarter of fiscal 2027, more than doubled year over year.
• This growth was driven by a surge in AI demand, leading to a global infrastructure build-out supported by diverse growth opportunities across various sectors.

Data Center Revenue Surge:

• Data center revenue reached $89 billion, with a 18% quarter-over-quarter increase.
• The growth was primarily driven by strong contributions from both hyperscale and ACINE segments, with notable demand for Blackwell and new GPU capacity.

AI Demand and Infrastructure Constraints:

• NVIDIA anticipates a 70% revenue growth in fiscal 2028, citing a supply-constrained outlook.
• The increase in AI demand is creating a significant need for infrastructure, with NVIDIA's compute being essential for various applications, from enterprise to sovereign AI projects.

Full-Stack AI Factory Platform:

• NVIDIA's full-stack AI factory platform, since Hopper, has expanded its share of the data center TAM, increasing revenue opportunity per gigawatt from $18 billion to $40 billion with Vera Rubin.
• This expansion is attributed to extreme co-design across GPU, CPU, and networking, enabling substantial performance gains and economic benefits for customers.

Memory Scarcity and Pricing:

• Memory scarcity is impacting NVIDIA's gross margins, with a significant price increase exceeding prior expectations.
• The price surge is a symptom of the same AI demand surge that is driving NVIDIA's growth, as memory is essential for the increasing complexity and scale of AI workloads.

Sentiment Analysis:

Overall Tone: Positive

• CEO stated, "We delivered another outstanding quarter with record revenue, operating income, and EPS." CFO noted, "We expect to grow revenue by approximately 70% in fiscal 2028. This is a supply-constrained outlook." CEO expressed confidence in demand and supply chain alignment, stating, "what we wanted to do is to be consistent with everybody... And the reason why that's important is because... everybody's putting a lot of resources at play."

Q&A:

• Question from Joseph Moore (Morgan Stanley): Great, thank you. I wonder if you could give us Color on the 70%. And what gives you the confidence to guide a full year out if you haven't been doing that? And then what's the gap between that amount of growth and the 100% demand growth? You know, what is the kind of key constraint that separates those numbers? And could you close those gaps over time?
  Response: Demand is driven by diverse, non-hyperscale segments (sovereign, neocloud, enterprise, startups) growing ~100% annually. Supply constraints (land/power/shell, construction, supply chain) limit ability to meet all demand; 70% is a confident deliverable based on current supply chain alignment, with efforts ongoing to increase capacity.

• Question from CJ Muse (Cantor Fitzgerald): There's tremendous investor focus on your inference market share. Can you speak to the evolving workloads you're seeing with the Gentic AI and how you see your share evolving here over time, particularly when you reflect on the growing value of the TAM you're seeing with each new full-stack generation, your expectation for greater growth from ACIE, and then also including Grok 3 LPX?
  Response: NVIDIA's full-stack platform handles the entire AI lifecycle (data prep, training, post-training, agentic inference) on one fungible architecture, increasing value per gigawatt from $18B (Hopper) to $40B (Vera Rubin). Grok 3 LPX offers high interactivity for specific services, but the core will be VLink72 systems.

• Question from Stacey Rasgon (Bernstein Research): So the 70% growth in fiscal 28... I was just wondering if you could talk us through the contributors here. Does that increase across the different products of Vira and CPUs and Grok and everything else? And also, you talked about your price increase that takes effect in Q1... What would it be if it wasn't constrained?
  Response: Growth is driven by both hyperscalers and non-hyperscale segments (ACIE). The unconstrained demand growth is significantly higher than 70%; increasing capacity is a priority. Each generation increases revenue per gigawatt ($18B to $40B) and productivity.

• Question from Vivek Arya (Bank of America Securities): When I just add up everything that's in the CFO commentary, I get to a number of about $500 billion or so, obviously, over the next several years... Secondly, if there is a specific cash part of that that we should think about in fiscal 28. And then... how are you balancing this dynamic where you want to invest a lot of the ecosystem, but part of that ecosystem wants to develop competitive solutions?
  Response: The $500B+ commitments include supply agreements critical for Vera Rubin production and demand visibility. NVIDIA's full-stack platform is fundamental for AI labs' global deployment; even if labs build custom chips, they will rely on NVIDIA's technology and ecosystem for broad deployment and economics.

• Question from Timothy Arcuri (UBS): There's a lot of talk about that open models could gain share... So how do you sort of put and take that? Do you see the rise of open models as being good for NVIDIA or ultimately negative?
  Response: Both closed and open models are skyrocketing and largely run on NVIDIA. Open models are vital for startups and enterprises to build proprietary AI, while closed models drive profitable token generation. NVIDIA benefits as long as models succeed, and its architecture runs every frontier model.

• Question from Joseph Moore (Morgan Stanley): You talked about demand growing 100% next year... with the developments in RSI, as well as AGI, what happens to industry demand? Does Does it inflect further? And what does it mean for NVIDIA when those things take place?
  Response: Demand will inflect further as AI becomes fully agentic, with agents running continuously in the background, driving recursive self-improvement and more profitable token generation. NVIDIA benefits as increased compute leads to more profit for services.

• Question from Jim Schneider (Goldman Sachs): Can we talk about some of the rank order, some of the most uh acute constraints whether that be things like uh data center power and shell availability uh d-ram wafer foundry uh availability etc if you can maybe help us understand we know which are the biggest among those that would be very helpful
  Response: The entire supply chain is challenged (land/power/shell, construction, components). NVIDIA has supply for ~70% of demand and is working with suppliers to increase capacity daily; transparency with partners is key to meeting customer needs.

• Question from Aaron Rakers (Wells Fargo): should we really conceptualize like 40 billion goes to 60 billion, 80 billion? And then... how do we kind of think about your ability to scale the capacity deployments?
  Response: Revenue per gigawatt will continue to increase with each generation (Hopper $18B → Grace Blackwell $25B → Vera Rubin $40B). The goal is to maximize compute per gigawatt/land, enabling faster ROI and driving further investment.

Contradiction Point 1

Growth Outlook for Non-Hyperscale (ACINE/ACIE) Segment

Contradiction on whether ACINE growth is the primary driver for exceeding hyperscaler CapEx growth.

Stacey Rasgon (Bernstein Research) - Stacey Rasgon (Bernstein Research)

2027Q2: The increase is driven by both hyperscaler and non-hyperscaler (ACINE) demand. - Jensen Wong(CEO)

Regarding the increased $200 billion growth outlook for fiscal 2028, what are the contributors across products, and what would growth be without constraints? - Ben Reitzes (Melius Research)

2027Q1: NVIDIA should grow faster than hyperscale CapEx because its data center business includes... 2) AI-native, enterprise, and sovereign data centers... This segment is growing incredibly fast and is expected to surpass hyperscale in size over the long term. - Jensen Huang(CEO)

Contradiction Point 2

Market Role and Growth Expectations for LPX Products

Contradiction on whether LPX is a minor niche product or a significant future component.

CJ Muse (Cantor Fitzgerald) - CJ Muse (Cantor Fitzgerald)

2027Q2: Grok 3 LPX offers high interactivity and low latency, catering to high-ASP services, while the majority of data centers will use Vera Rubin and MVLink 72. - Jensen Wong(CEO)

How are generative AI workloads evolving, and how is your market share changing with the growing TAM and Grok 3 LPX? - Timothy Arcuri (UBS)

2027Q1: LPX (like Rubin CPX) is a niche product designed for low-latency, high-token-rate services... NVIDIA expects LPX to remain a small part of the market for the foreseeable future. - Jensen Huang(CEO)

Contradiction Point 3

Supply Constraints and Growth Confidence

Contradiction on the severity and impact of supply constraints on growth guidance.

Joseph Moore (Morgan Stanley) - Joseph Moore (Morgan Stanley)

2027Q2: The supply constraint is a bottleneck, but NVIDIA is working with its supply chain to increase capacity. The demand is much greater than 70%, but the supply allows for confident delivery of 70%. - Jensen Wong(CEO)

What factors support the 70% growth guidance for the full year? - Benjamin Reitzes (Melius Research)

20260226-2026 Q4: We are managing supply chain constraints and have built a significant buffer of inventory to ensure we can meet customer demand. - Jen-Hsun Huang(CEO)

Contradiction Point 4

Revenue Opportunity per Gigawatt

Contradiction on the magnitude of the revenue opportunity per gigawatt of data center.

CJ Muse (Cantor Fitzgerald) - CJ Muse (Cantor Fitzgerald)

2027Q2: Each gigawatt of data center represents a much larger revenue opportunity ($40 billion with Vera Rubin). - Jensen Wong(CEO)

How are generative AI workloads evolving, and what impact is this having on your market share? - Stacy Rasgon (Bernstein Research)

20260226-2026 Q4: We are seeing a massive sequential acceleration from Blackwell to Rubin, similar to what we saw with Blackwell. - Colette Kress(CFO)

Contradiction Point 5

Supply Chain Constraints and Growth Outlook

Contradiction on the severity and nature of supply constraints affecting growth projections.

Jim Schneider (Goldman Sachs) - Jim Schneider (Goldman Sachs)

20260226-2026 Q4: The exact ramp rate is still being determined. - Colette Kress(CFO)

What are the key supply constraints impacting the 70% growth outlook? - Christopher Muse (Cantor Fitzgerald)

2027Q2: The entire supply chain is challenged, including data center power/shell availability, DRAM wafers, and foundry capacity. - Jensen Wong(CEO)