Optimisus
Live markets
BTC—— ETH—— SOL—— BNB—— XRP—— DOGE——
View all markets
CategoryTechnology

Nvidia’s Data Center Segment Booked $89 Billion in a Single Quarter

By Optimisus • •4 min read
Nvidia’s Data Center Segment Booked $89 Billion in a Single Quarter

Nvidia reported fiscal second-quarter 2027 results after the close on August 26.

Data center revenue reached a company record of $89.02 billion, up 116.6% year over year and 18.3% quarter over quarter.

The stock rose 8.7% the following day, its biggest single-day percentage gain since April 2025.

The Segment Breakdown

Chief financial officer Colette Kress attributed the growth to the ramp of Blackwell Ultra infrastructure.

Hyperscale revenue rose 101.5% year over year and 13.1% quarter over quarter to $48.71 billion.

The category Nvidia labels ACIE, covering AI clouds, industrial and enterprise, grew 138.1% year over year and 25.2% quarter over quarter to $40.31 billion. Kress described that as driven by end demand from AI natives, enterprises and sovereign customers, plus hyperscalers using AI clouds.

That second category growing faster than hyperscale is the more interesting detail. It suggests demand broadening beyond a handful of very large buyers.

Read the Analyst Estimates Carefully

Pre-earnings coverage produced a wide and partly incoherent range of expectations, which is worth flagging for anyone reading back through it.

Several previews cited consensus revenue between $93 billion and $95 billion for the quarter, representing roughly 95% to 96% year-over-year growth.

At least one preview cited a figure of $3 billion to $5 billion, which is inconsistent with every other estimate and appears to be an error rather than a different methodology.

The actual data center figure of $89.02 billion is a segment number, not total company revenue. Comparisons between segment results and total-company estimates are a common source of confusion in coverage of this company.

The Macro Context It Landed In

The report arrived against a 10-year Treasury yield above 4.7%, a level not seen in years, which normally pressures growth valuations.

It also arrived during a week when Federal Reserve Chair Kevin Warsh delivered a hawkish Jackson Hole speech that flipped rate-cut expectations from near-certain to uncertain.

Nvidia rose anyway. That divergence between soft consumer data and technology earnings is the defining market pattern of 2026, and it is why AI infrastructure is being treated as somewhat independent of the broader cycle.

Whether that independence survives a sustained rate environment is the open question, and one quarter does not answer it.

The Rotation Underneath

One statistic complicates the bull case. The PHLX Semiconductor Sector index gained 63% during 2026 against roughly 21% for Nvidia’s own stock.

Investors have been rotating into other semiconductor names, which suggests the market views Nvidia’s dominance as priced and is looking for value elsewhere in the supply chain.

Nvidia’s stock generated a total return of 14.9% in the second quarter against the S&P 500’s 15.2%. The bellwether slightly underperformed the index it is supposed to be leading.

Hedge fund positioning was net constructive between March and June, with 86 funds initiating new positions and 417 increasing stakes, against 69 closing and 354 decreasing.

Why Crypto Readers Should Care

Two direct connections.

Bitcoin miners have spent 2026 converting contracted power into AI hosting revenue, a shift Optimisus documented in the piece on listed miners switching off rigs faster than the network is shrinking. Their revenue now tracks GPU demand.

Compute is also becoming a tradeable commodity. The CFTC opened a comment window on compute derivatives, and CME has planned futures on H100 and B200 rental indices, covered in the piece on the offensive security model released to vetted defenders.

A quarter like this sets the reference price for both.

What to Watch Next

Guidance rather than results. Nvidia’s forward quarter guidance and management commentary on Blackwell ramp progress determine whether the buildout is in expansion or digestion.

Model efficiency cuts the other way on demand, a tension covered in the piece on falling active parameter counts.

Hyperscaler capital expenditure is the leading indicator. Collective 2026 spending has been projected above $600 billion, much of it tied to AI infrastructure, and that trajectory underpins the growth case more than any single quarter does.

Sources

This is not financial advice.

Optimisus covers crypto and technology news for readers who want the detail behind the headline.

Topics
#AI Infrastructure #Blackwell #Data Centers #Earnings #NVIDIA #Semiconductors