As Nvidia prepares to report fiscal second-quarter earnings on August 26, Wall Street is bracing for another beat, but UBS argues the more critical test may come a quarter later. The chipmaker's guidance of roughly $91 billion for the July quarter is widely expected to be exceeded, yet the real question is whether the company can sustain its breakneck growth as it transitions to new architectures.
UBS sees Q3 revenue above $110 billion
UBS analyst Timothy Arcuri projects Nvidia will deliver $94 billion to $95 billion in revenue for the July quarter, followed by guidance of $107 billion to $108 billion for the October quarter. More strikingly, Arcuri believes actual third-quarter revenue could exceed $110 billion as demand for Blackwell remains firm and the first Rubin systems begin contributing to sales.
According to TipRanks, Arcuri expects Blackwell demand to stay steady while Rubin starts to ramp ahead of a larger acceleration later in the year. UBS has raised its calendar 2027 revenue forecast to $681 billion from $649 billion and lifted its GPU shipment estimate to about 10.8 million units from 9.2 million.
A quarter above $110 billion would not just mark another record; it would signal that Nvidia can move between major architectures without the revenue pause investors sometimes fear during product transitions. This is particularly relevant as the company navigates the shift from Blackwell to Rubin.
Rubin could extend the upgrade cycle
Bank of America sees a similar setup. Analyst Vivek Arya expects Nvidia to report $94 billion to $95 billion of second-quarter revenue and guide to $107 billion to $108 billion for the third quarter. He believes Vera Rubin shipments, new Vera CPU products, and continued cloud spending could trigger a "multi-quarter upgrade cycle."
The economics of the new systems could be equally important. BofA estimates Vera Rubin NVL racks could cost roughly $7 million to $8.5 million, compared with around $4 million for Blackwell Ultra. That higher system value could help Nvidia absorb rising memory costs while protecting profitability. BofA expects long-term gross margins around 73% to 74%, even after accounting for higher memory costs.
That means Rubin does not need explosive unit growth alone to expand Nvidia's revenue opportunity. Higher-value systems can also push sales higher, a dynamic that could be crucial as memory costs rise.
Financing becomes the next test
The larger question is whether the AI spending machine supporting those forecasts remains credible. Nvidia recently announced partnerships with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR to establish independent compute-financing platforms designed to mobilize more than $500 billion of third-party capital for AI infrastructure. This move, as Nvidia's $500B AI funding plan puts leverage in the spotlight, aims to ease concerns about circular financing.
Morgan Stanley analyst Joseph Moore said the structure "should arguably alleviate circularity concerns," according to MarketWatch, because sophisticated outside investors would provide most of the capital. Bank of America's Arya called the arrangement a "structurally bullish setup," arguing that it strengthens Nvidia's CUDA ecosystem while shifting much of the financing burden away from Nvidia.
There is still a catch. Moore described Nvidia's financial backstops for some neocloud and sovereign-AI customers as "the next big debate for the stock." That debate matters because Nvidia's revenue forecasts ultimately depend on customers continuing to fund enormous infrastructure programs and earning acceptable returns on them. The August 26 report is therefore about more than whether Nvidia beats its $91 billion outlook.
As the company's AI financing plan eases circularity fears, investors will be watching for signs that the demand curve remains intact. With AI memory opportunities expanding, the stakes are high for Nvidia to deliver not just a beat, but a credible path forward.
This article is for informational purposes only and does not constitute financial advice.
