Advanced Micro Devices (AMD) reported third quarter earnings yesterday, October 24, of 10 cents a share (excluding one-time items). That was better than the Wall Street consensus of 8 cents for the quarter. Revenue grew 257% to $1.64 billion in the quarter, surpassing Wall Street’s projection of $1.51 billion in revenue.
But while AMD raised its guidance for the fourth quarter, it still projected a 12% to 18% sequential drop in fourth quarter revenue. That was right on the mark for revenue with the company’s $1.34 billion to $1.45 billion projection hitting Wall Street’s estimate of $1.34 billion.
The drop in fourth quarter revenue would be the first sequential decline in quarterly revenue in seven quarters. (Year over year the company guided to a 26% increase in revenue for the fourth quarter.)
The stock, which closed up 1.06% before the earnings announcement dropped 10.95% in after-hours trading on the news.
Before AMD’s earnings report traders had speculated that the company’s earnings would be a useful indicator for what Nvidia (NVDA) would report on November 9. After all, the two companies do compete in some of the same markets with processors for gaming, and servers, and, increasingly for chips with parallel architectures used in artificial intelligence and autonomous driving applications.
Sales for AMD’s graphics and computer business, which makes processors for desktop and mobile PCs, servers and game players such as Microsoft’s Xbox, were up 73.5% to $819 million. (A year ago this segment lost $66 million. This year it came in $70 million in the black.)
The problem for AMD, and perhaps by extension for Nvidia, is that the third quarter typically marks a seasonal peak in revenue for AMD and many other chipmakers because customers in the computing and gaming industry put in big orders in the third quarter so they’ll have product to sell in the fourth quarter.
The hope is, always, that a chip maker will have found a market that doesn’t swing quite so violently with the shift from third to fourth quarters. That hope is frequently (or almost always) dashed, but Wall Street persists in treating the cyclical decline in revenue in the fourth quarter as big negative news.
One problem that AMD has, and Nvidia doesn’t, is the the company doesn’t break out segment sales (Nvidia does) so investors really don’t have any idea how much penetration AMD has achieved in any of the fast growing, new markets that have so excited traders in Nvidia–and to a lesser but still significant extent in AMD and Intel (INTC). The company said that the strong year to year growth in computing and graphics revenue was primarily due to strong sales of Radeon graphics processors. And AMD announced some splashy deployments with companies such as Amazon and Baidu, but it was hard to tell from either the slides or the conference call, exactly how much traction the newest generation of graphic processors in the Radeon family had gained. Same goes for the company’s data center effort centered on the EPYC chip. It is clearly being tested by potential customers but I couldn’t get a sense of adoption rate for EPYC or the new Ryzen processor and its mobile version. In response to a direct question on the growth rate for these new products, the company simply affirmed that sales were ramping.
In after-hours trading yesterday the market decided that AMD’s results didn’t have a lot of significance for Nvidia. A lot of the selling in AMD was typical sell on the fact as traders who had bought hoping for news that Tesla or some other big name had adopted AMD chips didn’t hear anything concrete. Nvidia shares were down just 0.44% in the after-hours session. The takeaway from that, though, is that traders are probably expecting some concrete news on the markets that everyone is excited about from Nvidia. If they don’t get it (even if the long-term prospects for those markets remain strong,) we could see an after-hours reaction on November 9 like that in AMD today