A sharp slowdown in demand from crypto-currency miners eases the enthusiasm for the in many ways strong Nvidia (NVDA) reporting report.
After the bell Thursday, Nvidia reported sales of $ 3.21 billion (+ 66% annually) and GAAP earnings per share of $ 1.98 in April (fiscal quarter). This exceeded the consensus analyst estimates of $ 2.89 billion and $ 1.46. All five of Nvidia's reported market segments – gaming, data center, automotive, professional visualization, and OEM & IP – outperformed consensus revenue estimates.
Nvidia also contributed $ 3.1 billion to Juliquartal, up or down 2%. This represents a 39% growth in the middle, above a $ 2.95 billion consensus. Nonetheless, post-trading equities fell 2.9% to $ 252.55. They sold after Chief Financial Officer Colette Kress announced on Nvidia's profit share that sales of products designed specifically for cryptocurrency miners would account for about one third of quarterly sales of $ 289 million in April (9% of total sales of Nvidia). GPU arch-rival AMD's (AMD) share fell 1
High expectations are a reason for the decline: Nvidia posted gains near an all-time high of $ 260.50. Here are some key findings from the Nvidia earnings report:
- Before the slowdown in the July quarter, Nvidia's crypto demand in the April quarter provided a healthy revenue boost. The sale of the above-mentioned cryptospecific products helped Nvidia's total sales of OEM and IP products (including non-gaming GPUs sold to PC OEMs) more than double sequentially and annually to $ 387 million. And with some buyers buying regular gaming GPUs, this was one reason (though by no means the only one) that Nvidia's gaming segment revenue grew 68% per year to $ 1.72 billion.
- Although an improvement in the graphics card was expected and recent statements by AMD and Taiwan Semiconductor (TSM), the magnitude of the expected decline in the July quarter seems to have surprised some investors. In addition to lower cryptocurrency prices and higher levels of mining complexity, the launch of an Ethereum Mining ASIC by Bitcoin mining ASIC developer Bitmain could be significant.
- GPU demand from gamers is strong, but only how strong is a little tricky. In addition to sales to miners, gaming revenue also benefited from Nvidia's efforts to replenish the graphics card channel inventories consumed by the miners, as well as rising Tegra processors for the Nintendo switch console. With graphics card accessories for games that are now much healthier, Nvidia will not get a similar boost in sales through channel replenishment this quarter. Combined with lower OEM and IP sales, Nvidia appears to be doing slightly lower than the April quarter for July quarterly sales. However, CEO Jensen Huang claims that there is still a lot of "stale player" to talk to.
- Not surprisingly, Nvidia's data center segment had an outstanding quarter, with revenue up 71% to $ 701 million. Of course, strong demand from cloud giants for Nvidia's Tesla V100 server GPU (widely used for AI training purposes) played a role. But the company also reported that the number of GPUs used for inference (running trained AI algorithms against real data and content) has more than doubled. Increasing enterprise demand for Tesla GPUs for handling high-performance computing (HPC) workloads also contributed to this. And Huang quoted Nvidia's DGX systems – used by companies that do not want to build their own AI training systems – now make "a few hundred million [dollars]" in annual sales.
- Nvidia's spending growth continues Gradual acceleration: The company's GAAP operating costs increased 30% annually to $ 773 million, up 28% in the January quarter. The main driver: R & D spending up 32% to $ 542 million.
- When Huang worries about the growing AI training competition, he now shows signs. He insisted that the Tesla V100 outperform GOOGL's recently announced third generation Tensor Processing Unit (TPU), which appears to be about twice as fast as the second-generation apple-to-apples based generation program V100 is a big advantage (TPUs are specially designed to run Google's TensorFlow Machine Learning Framework). He also sounded unconcerned about efforts (like that of Intel (INTC) and various startups) to build programmable AI training processors. "[I] It's going to be tough, it's going to be expensive, and we've been doing it for a very long time," he said.
- Huang and Kress both talked about the effects of "Battle Royale" multiplayer games, such as Fortnite and PUBG on the gaming GPU market – both in terms on the acquisition of new players as well as the motivation of existing players to buy new GPUs. Huang also predicted that Nvidia's RTX ray-tracing graphics rendering technology, powered by a new Nvidia Workstation GPU, will boost the gaming industry by helping developers create high-quality game content. The comments precede the anticipated launch of the first 2018 gaming GPUs based on Nvidia's Turing architecture.
Eric Jhonsa of TheStreet previously reported on Nvidia's earnings report and a live blog
Jim Cramer and the AAP team have positions in Nvidia and Alphabet for her Action Alerts PLUS Charitable Trust Portfolio . Would you like to be warned before Cramer buys or sells NVDA or GOOGL? Learn more now .