Citi lifts Nvidia's price target on booming sovereign AI demand
The firm raised its price target on the chipmaker to $190 per share, implying a roughly 15% upside from Nvidia's current trading levels. Citi analysts said they see Nvidia capturing a larger piece of an expanding total addressable market (TAM) for data center infrastructure. The upgrade comes as Nvidia continues its charge toward a $4 trillion market cap, with shares up 12% in the past month.
"We believe sovereign demand is already contributing up to billions of dollars in 2025" and should ramp up further in 2026, analysts Atif Malik and Papa Sylla wrote. Nvidia is involved in "essentially every sovereign deal," the note said, making the company central to the global race to build national AI infrastructure.
The firm raised its 2028 AI compute TAM estimate to $563 billion, up 13% from $500 billion, and networking TAM to $119 billion, up from $90 billion, expecting sharp sales increases for Nvidia.
Citi noted that at Nvidia's recent Generative AI conference, participants discussed a possible benchmark for AI infrastructure: one supercomputer or 10,000 GPUs per 100,000 employees, a ratio that could drive massive enterprise and government buildouts.
Nvidia, whose Blackwell GB200 chips power many of these AI clusters, is already seeing accelerating deployment, according to Citi. Concerns about potential bottlenecks in Nvidia's supply chain have also eased, with Citi reporting that rack buildouts are happening "at a rapid pace." As the company prepares for its next-gen GB300 chips, analysts expect a smooth transition, crediting lessons learned from earlier platform shifts.
Citi now expects Nvidia's data center revenue to grow 5% in FY 2027 and 11% in FY 2028. Networking sales are projected to surge by 12% and 27%, respectively. This represents a 20% attach rate, indicating a rising demand for high-performance systems that link large AI clusters. Gross margins are also forecast to continue expanding, normalizing in the mid-70% range by year-end.
Still, the bank flagged downside risks, including renewed export restrictions under a potential second Trump administration. Bloomberg recently reported that Malaysia and Thailand could face scrutiny for suspected shipments to China.
For now, though, the AI gold rush — especially from public sector buyers — shows no signs of slowing.
"Nvidia has line of sight to tens of gigawatts of sovereign and enterprise AI factory buildouts over the next few years," Malik and Sylla wrote.
Francisco Velasquez is a reporter for Yahoo Finance. He can be reached on LinkedIn and X.

Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles


Bloomberg
20 minutes ago
- Bloomberg
Citi's Chronert Says S&P 500 Could Reach 7,000
Open Interest Scott Chronert, Citi US equity strategist, makes the bullish case for US equities during an interview on "Bloomberg Open Interest." (Source: Bloomberg)
Yahoo
38 minutes ago
- Yahoo
NVDA: Huawei Challenges Nvidia with New AI Compute Platform
July 28 - Huawei Technologies took the stage at Shanghai's World Artificial Intelligence Conference to reveal its CloudMatrix 384 AI computing system, a platform Reuters reports rivals Nvidia's (NASDAQ:NVDA) most advanced offering. Warning! GuruFocus has detected 5 Warning Signs with NVDA. Built on custom AI?optimized chips and high?speed interconnects, CloudMatrix 384 delivers multi?petaflop performance for large?scale model training. Huawei says the system taps innovations in chip design, cooling and power efficiency to match or exceed the throughput of Nvidia's H100 GPUs, which dominate data?center AI workloads. Analysts note that while Nvidia (NASDAQ:NVDA) holds roughly 80% of the high?end AI accelerator market, competitors like Huawei and Cerebras are closing the gap. CloudMatrix 384 could appeal to Chinese cloud providers and enterprises seeking alternatives amid U.S. export restrictions on Nvidia chips. Investors remain bullish on the company's next?generation Blackwell architecture. For Huawei, the launch signals an intensified push into global AI infrastructure, setting up a fierce technology duel between the U.S. and Chinese AI champions. This article first appeared on GuruFocus.
Yahoo
38 minutes ago
- Yahoo
Huawei's AI System May Pressure Nvidia Dominance
Huawei surprised many at the Shanghai AI conference when it rolled out its CloudMatrix 384 system. Instead of a handful of high?power chips, Huawei went for 384 of its 910C processors and stitched them together with what it calls supernode architecture. Warning! GuruFocus has detected 5 Warning Signs with NVDA. The upshot? On some benchmarks it lines up pretty closely with Nvidia's (NASDAQ:NVDA) GB200 NVL72, according to SemiAnalysis founder Dylan Patel. We heard from Huawei Cloud's Zhang Pingan in June that CloudMatrix 384 is already humming away on their public cloud. That's a big deal because it shows Huawei isn't just talking hardware it's actually running real?world workloads. Sure, Nvidia's B200 chips still boast higher efficiency per chip. But Huawei's play is all about scale and system smarts rather than raw chip muscle. If Huawei can deliver on price and availability, Nvidia could feel some heat especially in markets where cost rules. Of course, export rules and customer comfort with Chinese tech remain question marks. For now, CloudMatrix 384 is China's boldest bid yet to challenge the AI?infrastructure crown. Watch pricing, performance reports and any new U.S. export moves. They'll tell us whether this is a fleeting headline or the start of a real shake?up. This article first appeared on GuruFocus. Error in retrieving data Sign in to access your portfolio Error in retrieving data Error in retrieving data Error in retrieving data Error in retrieving data