NVIDIA splits reporting as chips fragment

Why it matters

Nvidia said on May 20 that it is changing how it reports its biggest business, splitting Data Center into Hyperscale and ACIE, a bucket that includes AI clouds, industrial customers and enterprises. The move came with first-quarter fiscal 2027 results showing total revenue of $81.6 billion and Data Center revenue of $75.2 billion for the quarter ended April 26. Nvidia said the new framework is meant to reflect its “current and future growth drivers.” The reporting change arrives as custom AI chips are gaining share with the largest cloud companies. TrendForce said in an October 2025 note that in-house ASIC shipments at cloud service providers are expected to grow 44.6% in 2026, compared with 16.1% for GPUs. Ben Thompson of Stratechery wrote on May 26 that Nvidia’s new disclosure draws a line between hyperscaler demand and the rest of the market. (nvidianews.nvidia.com) ### Why did Nvidia redraw its biggest segment now? Nvidia said on May 20 that Data Center will now contain two sub-markets: Hyperscale and ACIE. The company said Hyperscale will include public clouds and the world’s largest consumer internet companies, while ACIE will cover AI Clouds, Industrial and Enterprise customers. (trendforce.com) That matters because Nvidia’s old presentation grouped together customers with different buying patterns. Under the previous framework, Nvidia reported Data Center compute revenue of $60.4 billion and networking revenue of $14.8 billion in the quarter, but those figures did not separate the biggest cloud buyers from the rest. Thompson wrote that the new split delineates hyperscaler sales from “everyone else.” (nvidianews.nvidia.com) ### What is changing at the hyperscalers? TrendForce said cloud providers’ in-house ASICs are expected to outgrow GPUs in 2026 shipment growth. Its projection did not say Nvidia’s position would “collapse overnight,” but said the contest is broadening beyond chip performance into interconnects, switches, software and ecosystem control. Those buyers have the scale to justify designing chips for narrower workloads, especially inference and internal services. (nvidianews.nvidia.com) TechTimes, citing TrendForce, reported on May 26 that custom AI chips are set to grow at nearly three times the rate of merchant GPUs in 2026. ### Does faster ASIC growth mean Nvidia is losing control? (trendforce.com) Jensen Huang said in Nvidia’s earnings release that the company is “uniquely positioned at the center of this transformation” because its platform runs “in every cloud” and scales from hyperscale data centers to the edge. Nvidia’s first-quarter results still showed Data Center revenue up 92% from a year earlier. (techtimes.com) TrendForce also said the shift toward ASICs changes the battleground rather than ending Nvidia’s lead. Its note said Nvidia’s response is extending beyond single-chip performance and into interconnects and the software layer, including opening NVLink for customer ASIC integration. ### Why does the split matter for investors? The new format gives investors a cleaner way to watch whether growth is being driven by a handful of cloud giants or by a broader set of customers. (nvidianews.nvidia.com) A deceleration in Hyperscale alongside steadier ACIE demand would point to different spending patterns than a broad slowdown across both groups; that is an inference from Nvidia’s new reporting structure and TrendForce’s forecast for faster ASIC growth at cloud providers. (trendforce.com) Nvidia has not yet published a full time series under the new categories in the materials reviewed here. The company said only that it is transitioning to the new framework and that future disclosures will better reflect its growth drivers. ### What should readers watch next? Nvidia’s investor relations site said the company hosted its first-quarter fiscal 2027 earnings call on May 20 for the quarter ended April 26. (nvidianews.nvidia.com) The next quarterly release and accompanying materials should show whether Nvidia begins providing Hyperscale and ACIE figures that let investors track the split directly. (investor.nvidia.com)

Key numbers

• Nvidia said on May 20 it will split Data Center reporting into Hyperscale and ACIE, separating cloud giants from enterprise, industrial and AI-cloud customers.
• TrendForce said in October 2025 that in-house ASIC shipments at cloud providers will grow 44.6% in 2026, versus 16.1% for GPUs.
• Nvidia said on May 20 that it is changing how it reports its biggest business, splitting Data Center into Hyperscale and ACIE, a bucket that includes AI clouds, industrial customers and enterprises.
• The move came with first-quarter fiscal 2027 results showing total revenue of $81.6 billion and Data Center revenue of $75.2 billion for the quarter ended April 26.

What happens next

• Nvidia said on May 20 that it is changing how it reports its biggest business, splitting Data Center into Hyperscale and ACIE, a bucket that includes AI clouds, industrial customers and enterprises.
• TrendForce said in an October 2025 note that in-house ASIC shipments at cloud service providers are expected to grow 44.6% in 2026, compared with 16.1% for GPUs.
• Ben Thompson of Stratechery wrote on May 26 that Nvidia’s new disclosure draws a line between hyperscaler demand and the rest of the market.

Sources

• nvidianews.nvidia.com
• trendforce.com
• techtimes.com
• investor.nvidia.com

Quick answers