Key Points
Morgan Stanley analyst Joe Moore elevated Nvidia to his “top pick” in March.
KeyBanc raised Nvidia’s target price to $300 from $275, expecting strong results driven by Blackwell Ultra and Rubin GPU demand.
Investors seek Nvidia’s strategy for AI inference amid competition, with CPU server racks expected at Computex in early June.
Nvidia stock has lost some momentum recently, but analysts believe the chip maker can reassert its dominance in artificial-intelligence processors when fiscal first-quarter earnings arrive this week.
Ahead of the report on Wednesday, shares were picking up fresh support on Wall Street. That included a fresh vote of confidence from Morgan Stanley analyst Joe Moore, who elevated Nvidia to his "top pick" in March. The price had fallen to a below-market valuation multiple at the time, creating what Moore saw as a buying opportunity.
It has taken time for investor enthusiasm to return to the story, as the secondary and tertiary beneficiaries of AI continue to outperform, Moore noted on Monday. (Optical networking companies like Lumentum and Coherent are just two examples of peripheral players that have rallied recently.)
Moore believes "the quarter will be a positive step towards a stock rerating." However, as demand for AI chips remains high and supply limited, Nvidia won't be able to say much on the earnings call to settle the longer-term debate over its ability to retain market share.
Investors should expect a "typical beat and raise pattern," Moore wrote, while raising his estimates for total data-center revenue. He rates the stock at Overweight with a $285 price target, up from $260.
Moore's bullish call came even as Nvidia shares slipped 1.3% to $222.23 on Monday, lagging behind the Nasdaq Composite. The stock's run-up ground to a halt last Friday as investors were rattled by rising borrowing costs.
On the heels of that pullback, analysts were largely upbeat. KeyBanc analyst John Vinh raised his target price to $300 from $275 and kept an Overweight rating on the shares, citing demand for Nvidia's graphics-processing units, which have cornered the market for high-performance tasks like AI model training.
"We expect Nvidia to report strong results and guidance driven primarily by increased demand for Blackwell Ultra GPUs and initial shipments of Rubin GPUs," Vinh wrote.
Investors are likely waiting to hear about Nvidia's plans to dominate inference -- the process of running AI models -- against the backdrop of ramping competition from the likes of newly listed Cerebras Systems.
Much of the excitement has centered on central-processing units (CPUs), which are viewed as the most cost-efficient way of conducting inference. Vinh expects Nvidia to announce stand-alone CPU server racks at the Computex conference in Taiwan next month.
D.A. Davidson analyst Gil Luria also raised his target price to $300, from $250. In his view, Nvidia's strong supply-chain management can be an advantage.
"Nvidia GPUs are much less of a bottleneck than CPUs, advanced packaging and memory, which isn't all a bad thing," Luria wrote. "Nvidia understood the demand patterns first, which is why it is adapting to memory shortages and even talking about the previously unthinkable -- selling CPUs on a stand-alone basis."
Paul Meeks, managing director at Freedom Capital Markets, noted that Nvidia is valued at roughly 20 times its expected earnings for fiscal 2028, with its adjusted earnings expected to grow 73% in fiscal 2027 and 36% in fiscal 2028.
"Unless you think Nvidia's business is going to collapse in the following year, which I don't, you should probably bet on this stock for the next 2 years, or at least up to $300," Meeks asserted.
