I remember when Nvidia (NVDA) lost the Apple (AAPL) MacBook, iMac, and Mac Pro sockets to rival Advanced Micro Devices (AMD). Going through old conference call transcripts, it is clear that when Nvidia lost Apple, analysts were deeply concerned, and pressed management on how it intended to win the Apple spot back.
Well, the reason Nvidia lost the spot was quite simple: "Fermi", Nvidia's last gen GPU, was a power hog that tried to do too much while AMD's chips at the time sipped far less power for similar performance. Apple went with the most power efficient chips at the time, and that honor fell to AMD. With "Kepler", Nvidia, obviously figuring out that its old strategy wasn't working, decided to bifurcate its graphics chip development.
For the mainstream gamers/consumers, it would produce a very power-efficient, high performance core that didn't try to worry about being double-precision beasts. Then for the high end, such as workstations, supercomputers, and so on, it would build a beefier, much more feature rich version of the chip. Not only did the more svelte, less power hungry chip suit many of its customers better, but the yields were better, too (i.e. gross margins were better).
"Kepler" won Apple, along with the majority of the notebook graphics share it had lost, back, as the mainstream "Kepler" took the crown as the most power efficient consumer GPU available.
With Apple, a vendor whose computers are strictly high end and spare no expense, on board, Nvidia is sure to see a year-over-year increase in the number of high end notebook-level graphics chips sold. However, there is a problem.
Apple's Mac Shortage And Nvidia
The sales may not be recognized as soon as investors would like. Apple actually has had a tough time getting its new iMacs shipped out. They're hard to make, so Apple, until recently, was quite supply constrained, and sold far fewer Macs during the most recent quarter than it would have liked. The pent up demand is likely to be satisfied during the current quarter, which may throw off traditional seasonality.
In the year ago period compared to the quarter that just ended on January 31, Nvidia did $953M in sales, and guided to $1.1B at the midpoint for this quarter. More interestingly, in the last fiscal year, the April quarter was down roughly 3%, in line with seasonality. Analysts are modeling a 2.7% sequential decline for the coming April quarter, following that traditional seasonal pattern.
However, there may be room for Nvidia to surprise on the upside in its guide, assuming the guide for the current quarter took into account Apple's supply shortage. As the new iMacs ramp (and these machines use high end, high margin parts), Nvidia could see a disruption in traditional seasonality for the year with a narrower-than-expected sequential decline in the guidance as a result.
It is unclear how material the Apple win is to Nvidia's top and bottom lines, but the mix of product sold to Apple is definitely on the higher end of the spectrum. If Apple accelerates its orders for high end iMac components as a result of the rest of the supply chain working without constraints, then this could mean some real upside from estimates for Nvidia in the April quarter. The only question is how much of this shortage was baked into the guide for the current quarter.
Disclosure: I am long NVDA.