Until April of this year, chip giant Intel's (INTC -10.14%) publicly stated goal was that it would begin manufacturing chips using its 10-nanometer chip technology in "high volume" during the second half of 2018. This schedule, however, was pushed out to sometime in 2019 because, according to Intel, the yield rate of the technology wasn't improving as quickly as the company had hoped.
In order to keep customers happy, Intel is continuing to release new products using upgraded versions of its 14-nanometer technology. These new products, though not as compelling as the 10-nanometer products that they're standing in for, deliver reasonable performance and power efficiency improvements over their predecessors.
Image source: Intel.
This strategy has been working well so far. In fact, last quarter Intel raised its full-year revenue guidance by $2.5 billion to $67.5 billion and, in tandem with its announcement that Brian Krzanich had resigned his post as CEO, the company announced that its second-quarter results blew away its original guidance. Don't be surprised if Intel raises its full-year revenue outlook again when it reports its full second-quarter results on July 26.
What's interesting, though, is that the 10-nanometer delay seems to have led to more demand than what Intel can currently handle for its 14-nanometer parts. Let's take a look at the evidence of that and what it means for Intel investors.
Two pieces of evidence
Back on May 8, DIGITIMES claimed that Intel had stopped supplying its lowest-end desktop processor platform controller hub (PCH), known as H310, to system vendors just a month after they launched.
A PCH is a kind of companion chip that handles several key features in a computer, such as USB connectivity and Serial ATA (SATA) connectivity for disk drives.
The reason that Intel reportedly stopped supplying the H310 chips was that Intel simply didn't have enough 14-nanometer manufacturing capacity to meet demand for the H310 chips.
"The sources said motherboard makers had been told that the supply of H310 would resume in July at the latest," DIGITIMES wrote.
Now, as of this writing in late June, there appear to be plenty of H310-based motherboards available, so whatever initial supply bottleneck that Intel had run into has apparently been resolved.
This isn't the only piece of evidence, though. A recent report from BenchLife.info, a website that regularly leaks out information about upcoming Intel products, says the company is considering canceling the introduction of what would have been its upcoming Z390 PCH for the high-end/enthusiast desktop processor market and instead relabeling its current Z370 PCH as Z390 in support of the company's upcoming ninth-generation Core processor launch for desktop computers.
The Z370 PCH is manufactured using Intel's older 22-nanometer manufacturing technology. Demand for Intel's 22-nanometer technology is set to fall off as Intel continues to transition its microprocessor and PCH products to 14 nanometer, so keeping the enthusiast desktop PCH chips on 22 nanometer could make sense, particularly as enthusiast desktop computers don't benefit from the increased integration and power efficiency of newer PCH chips as much as other types of personal computers do.
What does the 10-nanometer delay have to do with it?
At this point, you might be asking yourself, "What does the 10-nanometer delay have to do with the supply tightness of Intel's 14-nanometer chips?"
The answer is simple. Since it had, presumably, planned to move a significant portion of its chip volume to 10 nanometer by the end of 2018, the company likely expected demand for its 14-nanometer chips to be lower than what it'll now be.
Further exacerbating the problem is the fact that for Intel to continue to build better products using its 14-nanometer technology, it needs to do things like add more cores, which makes the chip sizes bigger. Bigger chip sizes mean fewer chips per wafer, which ultimately means that more wafers need to be produced for a given target shipment quantity.
Intel's solution
Intel has indicated that it doesn't expect to go into mass production on products using its 10-nanometer technology until sometime in 2019 (even then, the company won't publicly commit to whether it expects volume production in the first half of 2019 or in the second half of 2019). This means that it will need to be able to meet increasing demand for 14-nanometer products for at least another year.
The solution is, frankly, quite simple: Intel's putting in more 14-nanometer manufacturing capacity.
During the company's earnings conference call back in January , CFO Bob Swan said that Intel had raised its full-year capital expenditure outlook by $2.5 billion due to higher projected logic chip demand. Swan indicated that this spending would go toward its 14-, 10-, and even its 7-nanometer technologies.
It would seem, then, that Intel is putting in additional 14-nanometer capacity to handle this situation. Intel's goal is, frankly, to make sure it can build enough chips to meet demand and, based on the company's increased financial outlook for the second quarter of 2018, it looks like it's doing a pretty good job of doing that.
