Nvidia on Wednesday delivered strong quarterly results to cap off its fiscal year, outdone only by the chip giant’s outlook for the current quarter, in a sign the AI boom continues apace. Revenue in the company’s fiscal 2026 fourth quarter increased 73% year over year to $68.13 billion, outpacing the $66.2 billion the Street was looking for, according to estimates compiled by data provider LSEG. Adjusted earnings per share (EPS) increased 82% to $1.62, also exceeding the consensus estimate of $1.53, per LSEG data. Shares of Nvidia were slightly lower in extended trading, down about 50 cents apiece to $195.35. The stock had traded over $200 a share shortly after the numbers were released. The subdued reaction isn’t a total surprise, even if we would prefer to see those initial gains hold. The massive post-earnings moves for Nvidia that became routine in the early days of the AI boom have long since passed. NVDA 1Y mountain Nvidia’s stock performance over the past 12 months. Bottom line Talk about a strong earnings report. The results were good but the guidance for the current quarter is truly remarkable. Those who are skeptical of the sustainability of the AI boom keep waiting for the party to stop. Nvidia made it clear the party rages on. Sure, Nvidia’s quarterly revenues were roughly $2 billion more than expected. Even more impressive, though, is that the team’s guidance for the current quarter came in more than $5 billion ahead of estimates — despite the fact the Street had time to factor in the hyperscaler’s massive capex budgets for the year. In other words, analysts were right to raise their April quarter estimates in recent weeks. They just didn’t raise them enough. Equally notable, CFO Colette Kress said on the earnings call that Nvidia expects sequential revenue growth throughout 2026, exceeding the $500 billion revenue opportunity for its Blackwell and Rubin generation chips that CEO Jensen Huang disclosed in the fall. Blackwell is its current chip family, with Rubin set to launch later this year. “We believe we have inventory and supply commitments in place to address future demand, including shipments, extending into calendar 2027,” Kress said. As expected, Nvidia was asked about soaring memory costs and whether that poses a threat to the company’s ability to maintain its impressive gross margins in the mid-70% range, a key level for investors ( as we discussed in our week ahead preview ). In response, Huang said the “single most important lever” that Nvidia can pull to protect its margins is delivering “generational leaps” in performance. Our reading of that answer: If Nvidia’s products are still the best game in town , it’s easier to pass on any rise in input costs without having to eat the hit to margins. Meanwhile, Kress also offered an encouraging detail on demand for Nvidia’s older data center AI chips, which we think should help settle a big point of contention over the depreciation cycles for its hardware. “With Nvidia infrastructure in high demand, even Hopper and much of the six-year-old Ampere based products are sold out in the cloud,” Kress said. In the fall, some of Nvidia’s most important customers were catching heat for lengthening depreciation times in what critics called financial engineering designed to inflate their earnings. It got the point that mainstream media outlets, from our colleagues at CNBC to The Wall Street Journal , published deep dives into a debate usually reserved for a room of accountants. We’re not knocking those who were skeptical of the lengthened depreciation schedules, but this is a reminder of why we preach the importance of listening to earnings calls and getting information from those with their boots on the ground. That’s not to say we will blindly anyone at their word – there is almost always a conflict to be aware of – but when companies spend money, it’s because they believe it will lead to positive returns, period. The debate around the lifespan of these chips remains relevant, even if it’s not garnered as much attention lately. Why? The fact that six-year old Ampere generation chips are still being used is a sign of confidence to Nvidia’s customers, especially the cloud providers who drive so much of its data center businesses. It means the cloud providers — both giants like Microsoft and Amazon, as well as neoclouds like CoreWeave — can buy with confidence, knowing that the chips purchased today can generate revenue for many years into the future, even as Nvidia releases new-and-improved products at an annual cadence. There’s been questions whether releasing new chips every year will make some customers decide to just “sit out” a cycle, leading to an air pocket in demand. But this helps us understand why we haven’t seen it yet. Sure, the older chips may not be suitable for the most cutting-edge AI initiatives, but not everything being done in the cloud requires that. We figure that will certainly remain the case going forward as today’s most advanced AI applications are tomorrow’s run of the mill workloads. Bottom line, it’s clear that the demand for Nvidia’s chips is only increasing as demand for AI ramps up. Companies have acknowledged that not having a plan for AI is akin to not having plan for a website in the early 2000s or a plan for mobile apps following the launch of the iPhone. To put a plan in place, you need to be thinking about compute and there can simply not be a conversation about compute without including Nvidia. While custom chips, such as those made by Club name Broadcom , will have their place for certain applications where volume permits it, we don’t see Nvidia’s status as the AI computing king changing anytime soon. Kress’ commentary on demand are very bullish for the year ahead and, more importantly, backed by publicized spending intentions. As a result, we’re reiterating our $230 price target and our hold-equivalent 2 rating. We’re still optimistic, but are simply looking for a better opportunity to upgrade. Commentary Data Center , by far the most important of Nvidia’s five operating segments, saw revenue growth accelerate to 75% year over year, coming in at to $62.3 billion, better than $60.7 billion the Street was anticipating. Within the data center unit, compute revenue rose 58% year over year to $51.3 billion, while networking revenue surged 263% to $10.98 billion. Hyperscale customers accounted for a little over 50% of the revenue here. However, we liked that Kress said in prepared remarks that “growth was led by the rest of our Data Center customers as revenue diversified.” While we acknowledge that few companies on the planet can match the spending power of the hyperscalers, a broadening demand base will be crucial to blunting the impact of any one large customer pulling back on spend in the future. Gaming saw revenue jump 47% year over year to $3.73 billion. However, it still came up short versus estimates of $4.03 billion. Growth was driven by demand for the companies new Blackwell architecture. On the release, the team noted that supply constraints are expected to be a headwind “in the first quarter of fiscal 2027 and beyond,” which can be chalked up to the memory shortage. Professional Visualization revenue advanced 159%, topping expectations thanks to “exceptional demand for Blackwell.” Meanwhile, Automotive revenue was up 6% year over year thanks to continued adoption of self-driving platforms. The OEM & Other segment saw revenue increase 73% versus the year ago period to $161 million. This unit at Nvidia covers partnerships with original equipment manufacturers, licensing, and other things not accounted for in the other segments. Guidance Looking ahead to the current fiscal 2027 first quarter, management’s outlook was well ahead of expectations. Revenue of $78 billion, plus or minus 2%, way ahead the $72.6 billion LSEG consensus estimate. Adjusted gross margins are expected to be 75%, plus or minus 50 basis points, which at the midpoint is better than the 74.5% estimate compiled by FactSet. Expectations for adjusted operating expenses in the fiscal first quarter of $7.5 billion. The team is not assuming any Chinese sales in this guide, so any progress in trade talks between now and the end of the quarter should amount to upside. (Jim Cramer’s Charitable Trust is long NVDA and AVGO. See here for a full list of the stocks.) As a subscriber to the CNBC Investing Club with Jim Cramer, you will receive a trade alert before Jim makes a trade. Jim waits 45 minutes after sending a trade alert before buying or selling a stock in his charitable trust’s portfolio. If Jim has talked about a stock on CNBC TV, he waits 72 hours after issuing the trade alert before executing the trade. THE ABOVE INVESTING CLUB INFORMATION IS SUBJECT TO OUR TERMS AND CONDITIONS AND PRIVACY POLICY , TOGETHER WITH OUR DISCLAIMER . NO FIDUCIARY OBLIGATION OR DUTY EXISTS, OR IS CREATED, BY VIRTUE OF YOUR RECEIPT OF ANY INFORMATION PROVIDED IN CONNECTION WITH THE INVESTING CLUB. NO SPECIFIC OUTCOME OR PROFIT IS GUARANTEED.
Add Comment