Not too long from now, chip giant Intel (INTC 0.06%) is expected to launch its eighth-generation Core processors for the personal computer market. These products are expected to be manufactured on a third generation of the company's 14-nanometer technology, known as 14-nanometer++, which promises to deliver performance and power improvements over its 14-nanometer+ technology that's being used to build the bulk of its products today.
What's also interesting to note is that with its eighth-generation Core chips, Intel is planning to increase the number of processor cores within its notebook and desktop personal computer processors.
Image source: Intel.
The Kaby Lake-U Refresh part -- which is expected to replace the company's current Kaby Lake-U processors for mainstream notebook personal computers -- is expected to include four processor cores, up from the two cores found in today's Kaby Lake-U.
For high-performance notebook as well as for mainstream desktop personal computers, Intel is expected to replace Kaby Lake-H -- a quad-core processor -- with Coffee Lake-H, which is a six-core processor.
For mainstream desktop personal computers, Intel is expected to launch chips with up to six cores, a bump from the maximum of four cores on the current mainstream desktop personal computer chips.
While the additional cores should help Intel improve performance and even create a potentially interesting selling point, there's a question worth asking: Will this increase in core count hurt Intel's gross profit margins?
What affects chip margins?
Many factors affect a product's cost and, ultimately, margin structures. However, for the sake of simplicity, there are two major factors that impact a chip company's gross profit margin on a chip.
The first is the chip manufacturing cost, which is itself essentially a function of the manufacturing technology used to build it, the complexity of the packaging technology, the yield rate of the chips, and the size of the chip.
The second is the chip selling price. The more a company can charge for a chip, the higher its margins at a given selling price.
As a quick example, if a chip costs $20 to make and it can be sold for $40, then the gross profit on the sale is $20 and the gross profit margin percentage is 50%. If a chip costs $30 to make and sells for $40, then the gross profit is just $10 and the gross profit margin is merely 25%.
Circling back to Intel's eighth-generation Core
So, with Intel moving to higher-core count processors with its eighth-generation Core processor family, we know that the chip sizes will go up a bit (based on a leak from BenchLife.info, the move from the quad-core Kaby Lake-H to the hex-core Coffee Lake-H should drive a roughly 18% increase in chip size).
Now, an increase in chip size should -- assuming all else equal -- mean higher chip costs. I suspect that Intel's 14-nanometer yield rates have gone up year over year, so that improvement (which, all else equal, would increase the number of salable chips produced on a silicon wafer) would partially offset the increased costs from the larger chip sizes.
However, I don't think that improved yields could completely offset the chip size increases.
If Intel winds up selling these larger, higher-core count chips for roughly the same prices that it sold its previous products for, then the company could see some level of gross profit margin erosion. However, if the addition of processor cores allows Intel to increase its average selling prices generation over generation, then the company shouldn't see much, if any, of a negative impact to its product margins from the generational product transition.
