Intel's (NASDAQ:INTC) financial results don't lie and the company's most recent quarter indicates that the company is having some very serious issues with the yield rates of its 14-nanometer manufacturing technology.
Indeed, the company said in its most recent CFO commentary that its margins were negatively affected by the shift of its products away from the older 22-nanometer manufacturing process toward its 14-nanometer technology. If Intel's 14-nanometer yields were where they were supposed to be, this wouldn't be the case.
If Intel doesn't fix the issues that seem to be plaguing its 14-nanometer manufacturing technology, then this could have a material impact on the company's gross profit margins during 2016. Here's why.
Intel's entire product line transitions to 14-nanometer next year
During the first half of 2015, a large portion of Intel's chip shipments was on 22-nanometer. In particular, practically all of its desktop processors, all of its server processors, and the majority of its high-performance laptop chips were built on that older technology.
However, now that Intel has launched its 14-nanometer Skylake processor family, it won't be long before all of the company's high-end and midrange PC processors are built on 14-nanometer.
Additionally, Intel has been aggressively pursuing the low-end of the notebook and desktop PC markets. Although the company has some 14-nanometer products out to address those segments today, there are still many systems shipping with Intel's older 22-nanometer-based Pentium and Celeron processors.
According to various leaks, Intel is planning to refresh its low-end PC processor lineup with a 14-nanometer product family known as Apollo Lake. Since these chips will be targeted at low-cost systems, the chips themselves must be relatively cheap to make.
It's not just PC processors that Intel will need to worry about on 14-nanometer next year, though. The company is expected to transition its Xeon E5 and E7 server processors to 14-nanometer early next year.
If Intel doesn't significantly improve its 14-nanometer yields ahead of the production ramp of these processors, then I see one of two possible outcomes. The first is that Intel prices these processors in a "business as usual" fashion but takes the margin hit until yields come under control.
The second, which is probably less likely in light of the fact that Intel is expected to see competition in this market heat up, is to pass the added production costs onto its data center customers by raising chip prices.
Will Intel be able to get its yields under control?
The most troubling thing about these yield issues is that Intel has been in mass production on a 14-nanometer product for over a year. And to make the situation worse, the start of mass production had already been pushed out by a quarter or so due to yield problems.
Intel CFO Stacy Smith indicated on the company's most recent earnings call that the company's 14-nanometer costs are coming down "pretty rapidly." However, in the CFO commentary, the company says that although its 14-nanometer unit costs should improve, a continued shift in mix of 14-nanometer products in the fourth quarter is going to take yet another toll on gross profit margins.
This is a topic that Intel needs to address during its upcoming investor meeting scheduled for Nov. 19. In particular, as an investor, I'd like to know when to expect 14-nanometer costs to come down to the point where products built on that process are as cost effective as the prior generation 22-nanometer parts.