NVIDIA FY2027 Revenue Breakdown: Data Center, Gaming & Auto Segments Explained

NVIDIA (NVDA) reported Q2 FY2027 revenue of $96 billion, more than doubling year-over-year for the fourth consecutive quarter of accelerating growth. Data center revenue reached $89 billion, representing 93% of total revenue, with the balance spread across Gaming, Professional Visualization, Automotive, and OEM segments. The composition of that $89 billion data center total, and how it is evolving, is the central driver of the near-term financial trajectory.

Data Center Sub-Segments: Hyperscale and ACIE

The data center segment grew 18% quarter-over-quarter to $89 billion in Q2 FY2027. Within that total, the two sub-segments moved at meaningfully different rates. Hyperscale revenue came in at $49 billion, up 13% sequentially, driven by continued Blackwell deployments across major cloud operators. Amazon Web Services is deploying an additional 2 million GPUs starting Q2 FY2027 through Q2 FY2029, illustrating the multi-year scope of hyperscaler infrastructure commitments. The top five hyperscalers are expected to spend nearly $800 billion in aggregate CapEx in 2026, rising to $1.3 trillion in 2027.

The ACIE sub-segment, which groups NeoCloud providers, sovereign AI programs, industrial deployments, and enterprise customers, reached $40 billion in Q2 FY2027. That represents 25% sequential growth and 138% year-over-year growth. Sovereign AI revenue alone grew 35% sequentially and more than tripled year-over-year. Management noted on the earnings call that ACIE is expected to represent roughly half of data center revenue over the long term, up from its current share of the $89 billion total.

NeoCloud partners are a significant driver of the ACIE acceleration. These operators exited 2025 with approximately 3 gigawatts of installed capacity and are expected to exit 2026 with 8 gigawatts, a roughly 2.7x expansion over a single calendar year. That rate of capacity growth reflects how quickly non-hyperscaler AI infrastructure is scaling. The Beta Finch [NVIDIA Q2 FY2027 earnings podcast](/podcasts/NVDA_Q2_2027) covers management commentary on NeoCloud deployment timelines in detail.

Supply-Constrained Growth: The FY2028 Outlook

NVIDIA guided Q3 FY2027 revenue to approximately $108 billion, plus or minus 2%. For FY2028, management guided to approximately 70% revenue growth, framing that figure explicitly as a supply-constrained outlook. Customer demand forecasts pointed to the possibility of doubling, meaning the approximately 70% figure reflects production capacity limits rather than a demand ceiling.

The hyperscaler CapEx figures contextualize the demand environment. Nearly $800 billion across the top five customers in 2026, rising to $1.3 trillion in 2027, represents capital commitment at a scale that exceeds what the current supply chain can absorb in a single fiscal year. The binding constraint on FY2028 revenue growth is component availability and production ramp, not customer willingness to commit capital.

Gross Margin Trajectory and Revenue per Gigawatt

Gross margin held at 75% in Q2 FY2027. Management guided the margin to bottom in Q4 FY2027 at 71% to 72% as Vera Rubin ramps and component costs weigh on the mix, with a recovery expected thereafter. The temporary compression reflects the cost profile of a new architecture launch rather than a structural change in pricing power.

The revenue-per-gigawatt metric frames the long-run economics across architecture generations. Hopper-based deployments generated approximately $18 billion of revenue per gigawatt of installed capacity. Blackwell raised that figure to $25 billion per gigawatt. Vera Rubin is expected to reach $40 billion per gigawatt. Each successive generation has expanded revenue density substantially, which means the margin compression accompanying Vera Rubin's ramp is occurring alongside a meaningful increase in revenue per unit of physical capacity.

NVIDIA described Vera Rubin as the fastest product ramp in the company's history. That framing aligns with the AWS deployment timeline spanning roughly eight fiscal quarters and the NeoCloud capacity build that is more than doubling over calendar year 2026.

Forward Signals and the Evolving Revenue Mix

The combination of Q3 guidance at $108 billion, an AWS GPU deployment covering roughly eight fiscal quarters, and NeoCloud capacity scaling from 3 to 8 gigawatts points to sustained volume growth through at least mid-FY2029. The supply-constrained characterization of the FY2028 outlook means revenue growth is paced by production, not demand; the demand signal itself is larger than the guided outcome.

The more structurally notable development is what the ACIE trajectory implies for revenue concentration. With hyperscale at $49 billion and ACIE at $40 billion in Q2, the gap between the two sub-segments has narrowed considerably. If ACIE converges toward 50% of data center revenue as management indicated, sovereign programs, NeoCloud operators, and enterprise workloads, each with independent budget cycles and deployment drivers, become co-equal contributors to total revenue. That changes the revenue mix narrative from near-total hyperscaler dependency to a more distributed demand base, one where a slowdown in any single cohort of cloud operators carries less concentrated risk.

For a full record of NVIDIA earnings quarters, the Beta Finch [NVDA earnings archive](/groups/NVDA) covers each quarter's results and management guidance in podcast form.

• Q2 FY2027 total revenue: $96 billion, more than doubled year-over-year, fourth consecutive quarter of accelerating growth
• Data center revenue: $89 billion, up 18% quarter-over-quarter, representing 93% of total revenue
• Hyperscale sub-segment: $49 billion, up 13% sequentially
• ACIE sub-segment: $40 billion, up 25% sequentially and 138% year-over-year
• Sovereign AI revenue: up 35% sequentially, more than tripled year-over-year
• Q3 FY2027 revenue guidance: approximately $108 billion (plus or minus 2%)
• FY2028 revenue growth outlook: approximately 70%, supply-constrained
• Q2 gross margin: 75%; Q4 FY2027 expected floor of 71-72% before recovery
• AWS deploying 2 million additional GPUs from Q2 FY2027 through Q2 FY2029
• NeoCloud installed capacity: approximately 3 gigawatts at end of 2025, expected at 8 gigawatts by end of 2026
• Revenue per gigawatt: $18 billion with Hopper, $25 billion with Blackwell, $40 billion with Vera Rubin