Last year, I wrote quite a bit about the yield challenges that Intel (NASDAQ:INTC) was facing with its 14-nanometer chip manufacturing technology. The company disclosed during its November 2015 investor meeting that yields of its 14-nanometer technology still hadn't caught up with the kinds of yield levels seen with its 22-nanometer technology (despite the company forecasting a "closing" of this gap in 1Q 2015 all the way back in mid-2014).
These yield challenges, Intel CFO Stacy Smith said in his presentation, would lead to the following:
- Higher product unit costs in 2016 relative to 2015 in the performance segment of the PC market (i.e. quad core Core i5/i7 chips aimed at desktops, high performance laptops, etc.)
- Higher product costs for server processors (since Intel's server volume was essentially all built on 22-nanometer in 2015)
- Lower costs in 2016 for mainstream and value PC processors, since these segments largely transitioned to 14-nanometer product in 2015 (very negatively impacting product costs) and manufacturing yields will be better in 2016 than in 2015.
After the company's fourth quarter of 2015 results and updated guidance, I think we can gain a little more insight into how the company's 14-nanometer yields are progressing. Let's take a closer look.
Digging into the company's gross profit margin outlook
Going into the fourth quarter, Intel had guided to a gross profit margin of around 62%. The actual result for the quarter came in at a cool 64.3%, which is near the top of the company's historical range.
In the company's CFO commentary, the company attributed this gross profit margin beat to the following factors:
The potential clue here is the first line in which Intel says that its platform unit costs came in lower than expected. There are a couple of potential explanations for this:
- Intel shipped more high-yielding 22-nanometer silicon during the quarter than it had expected to, or
- Intel's yields on its 14-nanometer process improved more quickly than the company had expected.
Interestingly, as part of the company's earnings release, Intel said that as of November, more than 50% of the company's client computing group volume had migrated to the 14-nanometer manufacturing node. Although this data point doesn't tell us much about Intel's expectations going into the quarter versus what actually materialized, I am inclined to believe that the lower unit costs were more a result of better-than-hoped 14-nanometer yield than a greater-than-anticipated mix of 22-nanometer product.
Looking at the 2016 gross margin guide
At its most recent investor meeting, Intel guided to a full-year gross profit margin of 63% on a non-GAAP basis (excluding charges related to the company's acquisition of Altera). This represents about a point up from previous guidance.
Now, about 0.4 points of that benefit came from the fact that Intel is lengthening its manufacturing equipment depreciation timeline, but even excluding that it looks like gross profit margins should be up slightly relative to prior guidance even with the new, lower revenue forecast (which could potentially lead to a loss of leverage).
Perhaps this slight increase is reflective of Intel raising its expectations for 14-nanometer yields?
At any rate, I think investors should keep an eye on Intel's actual gross profit margin results for the year. I wouldn't be surprised if there were some conservatism baked into that gross profit margin guide vis-a-vis yields, or perhaps the company now expects any good news on 14-nanometer to be offset by other factors.