Nvidia reports August 26. The stock has had a rough few weeks. Bond yields spiked, the broader market sold off, and AI infrastructure names got hit harder than most.
Michael Burry flagged a startup. The circular financing debate keeps resurfacing. There is no shortage of reasons to be cautious heading into next week.
Oppenheimer is not cautious. The firm just reiterated its bullish case with numbers specific enough to be worth examining before the report lands.
Oppenheimer maintained its Outperform rating and $265 price target on Nvidia on August 20, according to Investing.com.
The stock trades at a P/E of 33.64 with a PEG ratio of 0.3. Nvidia has delivered 71% revenue growth over the past 12 months. Market cap sits at $5.31 trillion. Gross profit margin is 74%.
Oppenheimer expects Nvidia to beat second-quarter estimates and deliver a stronger-than-expected third-quarter outlook. The firm projects more than $1 trillion in cumulative revenues from Nvidia's GB200, GB300 and VR200 platforms between 2025 and 2027.
That is a platform-level projection for specific products. It reflects the scale of what hyperscalers and AI developers are committing to spend on infrastructure.
Nvidia just made a move Wall Street wasn't ready for
Nvidia just locked down deal that changes AI race
Nvidia stock is doing something it hasn't done in years
Nvidia guided to $91 billion in Q2 revenue, plus or minus 2%, with gross margin targets of approximately 75%. Wall Street's average estimate sits slightly above that at $91.9 billion. The guidance excludes any data center compute revenue from China.
The firm also noted Nvidia's commitment to SB Energy's PORTS-Pike Technology Campus in Pike County, Ohio. Nvidia will be the exclusive AI compute infrastructure provider at the campus, with initial capacity of 4.25 gigawatts and an option to expand to 8 gigawatts. OpenAI will lease the campus for 20 years. Nvidia is investing $1.5 billion directly in SB Energy and has provided residual value guarantees capped at $105 billion. Capacity comes online in phases beginning in 2028, CNBC reported.
Oppenheimer's near-term case rests on Blackwell Ultra. The next-generation VR200 system is expected to ramp during the current quarter, adding momentum to Nvidia's data center business in the second half of the year.
Blackwell Ultra's performance-per-watt advantage is central to the firm's argument. AI data centers consume enormous amounts of electricity. A chip that produces more computing performance per watt reduces operating costs and helps customers manage limited power capacity.
The firm highlighted two metrics it expects to become increasingly important as AI moves into commercial deployment. Tokens per minute, which measures how quickly a model generates output. Cost per token, which measures how expensive that output is. Oppenheimer says Nvidia is best in class on both.
