Investing.com — Nvidia is anticipated to report stronger-than-anticipated quarterly gross sales subsequent week, with Citi elevating its income forecasts on improved shipments of AI networking elements and a continued ramp-up of the corporate’s Blackwell chips.
Citi expects Nvidia to publish July-quarter income of about $93 billion, roughly $1 billion above the Wall Avenue estimate, helped by a robust ramp of its B300 methods. It forecasts October-quarter income of $105 billion, about $1.5 billion above consensus, representing a 13% sequential improve.
Citi maintained its Purchase ranking and $300 worth goal, whereas decreasing its valuation a number of assumption to 24 instances projected calendar 2027 earnings from 28 instances to replicate decrease broader market multiples.
The brokerage expects data-center income to rise 15% sequentially within the July quarter and 14% in October, in contrast with Avenue expectations of 13% progress in each durations. It mentioned faster-than-expected shipments of 1.6-terabit transceivers additionally level to an preliminary ramp of Nvidia’s next-generation Vera Rubin platform.
Citi mentioned it believes Nvidia has secured enough high-bandwidth reminiscence provide for 2026 and 2027, doubtlessly permitting analyst estimates to maneuver greater. It raised its fiscal 2027, 2028 and 2029 adjusted earnings-per-share estimates by 1%, 2% and a couple of%, respectively.
The brokerage additionally highlighted Nvidia’s rising function in financing AI infrastructure, pointing to its assist for the Ohio PORTS-Pike campus, which entails land, energy and shell development for an preliminary 4.25 gigawatts of IT capability. Citi mentioned the association may assist AI labs and cloud suppliers ease near- to medium-term financing constraints as they increase computing capability.
Citi expects Blackwell GPU shipments in fiscal 2027 to succeed in 7.7 million models, pushed largely by the B300 ramp, whereas decreasing its Rubin estimate to 2 million models from 2.2 million due to tighter reminiscence availability. It raised its fiscal 2028 whole GPU unit estimate by 7%, reflecting a shift towards extra servers with decrease reminiscence per chip.
