Nvidia just reported its Q1 FY27 earnings that trounced all expectations. The stock has staged quite a rally since April, up 33% since then, helping the Semiconductor ETF SMH create higher highs.
Nvidia $NVDA reported its highly anticipated Q1 FY27 earnings results yesterday that trounced all expectations.
Unpredictability from Nvidia’s China revenues returned. But excluding China, Nvidia’s AI business is powering through all the geopolitical headlines and supply chain distortions. Nvidia’s Blackwell and Rubin ramps are in full force but Jensen Huang wants everybody to know that their CPU business is going well too.
This is why Nvidia blindsided markets with radical changes to its segment revenue reporting that enforces the view that Nvidia should not be known as a GPU company anymore. It is an AI platform company.
Nvidia’s Q1 report carried a large cherry on top for large shareholders who had been demanding more focus on shareholder value over the past few months, as we previously reported. That should keep large institutional shareholders happy for now.
Nvidia Ups The Expectations Game
Nvidia’s Q1 FY27 results delivered record revenue of $81.6B, beating the consensus estimate of $78.9B by ~3.5%. Q1 revenues accelerated both y/y by 85% and sequentially by ~20%, which is the type of growth acceleration markets favor.
Nvidia’s explosive Q1 growth was primarily fueled by its DC (Data Center) division, which generated a record $75.2B, accelerating by 92% y/y. The expansion of networking hardware and the highly anticipated Blackwell architecture ramp were major drivers of this momentum.
Profitability metrics were strong ; the chipmaker posted gross margins of 74.9% (vs. 75% in Q4 and 60.5% last year), demonstrating robust pricing power despite the transition to newer architectures. On the bottom line, the $1.87 adj. EPS represented a massive 140% YoY jump or a 15% increase over Q4 FY26, also beating expectations of $1.76 per share.
Mind you, these numbers were delivered by Nvidia, with its China segment coming in weak.
Expect Nothing From China, Period.
Regarding China, Nvidia’s revenues from the region fell back closer to their longer-term quarterly dollar average of $3.5-4B. In Q1, China revenues fell 16.4% YoY to $4.6B, almost getting sliced in half versus their prior quarter’s China revenues of $8.4B.
Note that the US government authorized Nvidia to sell H200 GPUs to China in January. The Chinese government has also officially granted Chinese enterprises to buy Nvidia’s H200 GPUs, at least on a high level. However, the Chinese government is pushing for domestic adoption of homegrown GPUs, which makes Nvidia’s dependence on the China market extremely unreliable. Moreover, the China business appears to be margin dilutive for Nvidia since the US government takes a 25% cut on every H200 sale Nvidia makes to China.
To conclude Historically, Nvidia has always segmented its revenues along the lines of the end markets it serves. While Gaming, Automotive, and Visualization were once growth markets during Nvidia’s existence, these end markets have made way for the Data Center, with Compute and Networking being the two pillars of the DC segment.
In Q1, Nvidia surprised investors with a shiny new revenue segment reporting structure that split the AI platform company’s consolidated revenues by the 2 key market platforms to which it sells its products and solutions. While the DC segment remains intact, all other product revenue segments are now consolidated into Edge Computing.
However, Nvidia has made internal changes to how it further breaks out its DC business. Instead of reporting its DC revenues as Compute or Networking, Nvidia will now break out its DC revenue segment as Hyperscale and ACIE (AI Clouds, Industrial, & Enterprise).
