Just recently, South Korean electronics giant Samsung (NASDAQOTH: SSNLF) announced a next-generation chip manufacturing technology that it dubs "8nm." The technology, which Samsung says was completed three months ahead of schedule, offers a modest 10% improvement in performance and area compared to its upcoming 10nm LPP technology (which, to my knowledge, has yet to be used to produce commercially available chips).
This announcement led one prominent tech site, Engadget, to declare that "Samsung leapfrogs Intel (INTC 2.20%) again with 8-nanometer chips."
Image source: Intel.
Intel routinely claims that its 14nm technology is similar in feature density to foundries' 10nm-class technologies and says that its 10nm technology is about a full generation ahead of competing 10nm technologies (which would certainly include Samsung's 8nm technology).
As I've said before, and I'll say again, Intel is shooting itself in the foot by not properly aligning the names of its technologies with the naming conventions that the rest of the industry players use.
It's all about branding
When I was at Intel's investor meeting earlier this year, I specifically asked an Intel executive about why the company doesn't simply rename its manufacturing technologies to more accurately reflect their competitiveness with other industry technologies.
The response I got was something along the lines of Intel wanting to be "intellectually honest" about its chip manufacturing technologies.
Here's why that doesn't make sense.
Chip manufacturing technology names like 14nm, 10nm, 7nm, and so on aren't indicative of anything intrinsic to the technology. These numbers don't refer to any specific physical dimension or characteristic; they are purely marketing names for the sake of the consumer.
Since these names are for marketing/brand building purposes, Intel is being intellectually dishonest by short-changing its technology.
Moreover, the company is actively destroying the value of its brand by giving the impression that its upcoming manufacturing technology is a generation behind what competitors will be shipping in roughly the same time frame.
There's just no reason for this.
How Intel gets out of this bind
Since Intel has yet to ship its first 10nm-based processors, it should just go ahead and rename the technology now. Intel's 10nm technology should be comparable in density to the 7nm technologies its competitors plan to put out, so it makes sense for Intel to rename its 10nm technology to 7nm.
This would be intellectually honest because it'd accurately reflect its competitiveness with other similar technologies and it would, at the very least, save Intel from embarrassing headlines about how it has been leapfrogged again.
I would advise Intel against taking this a step further by calling its future 10nm+ and 10nm++ technologies by names such as 8nm because, at that point, it would start to be dishonest, especially since Intel has explicitly said that those future technologies improve performance and not density.
It might make sense for Intel to start delivering modest density improvements on top of its performance improvements for future technologies, though, to stay competitive. These improvements don't need to be large, but density improvements coupled with performance enhancements could allow Intel to build better products and to have a more attractive set of offerings to potential contract chip manufacturing customers.
