Spark Global Limited reports:
Intel recently released a new CPU process roadmap, revealing a number of new processes for its future, including the first Intel 20A and 18A for Emiprocess. But note that Intel’s biggest move this time around is to change its vests, rebranding its own processes to match the latest Samsung/TSMC processes.
Specifically, the 10nm Enhanced SuperFin is being renamed Intel 7,7 nm is being renamed Intel 4 and, in the future, Intel 3, which is enough to keep Intel from being outsmarted when it comes to TSMC’s 5nm and 3nm processes, at least in name.
In response to the change, Intel said that the previous process node naming rule, which dates back to 1997, was defined based on the actual gate width (XXNM), but in recent years, the process naming and the gate width have not matched, so Intel has adopted a new naming method to make it easier for industry, customers and consumers to understand.
It would be interesting to see why Intel, which has been sticking with the system for the past decade or so, chose not to use it this year. Intel doesn’t want to lose out.
In a recent conversation with Der8auer, Intel executives revealed that the nanomedicts based on the length of the grid have been around for 12 years, and they don’t reflect the current state of the art. Intel urged the industry to come up with a standard that all customers could understand, but Intel’s friends ignored the idea for years.
As for who Intel is talking about, although it is not named directly, everyone knows that it is Samsung/TSMC. Since the 28nm process, they have started to follow their own standards in naming their processes. Rumors about water injection have been circulating for years.
If you look at the density of transistors, Intel’s previous name does suffer. The density of 10nm technology is as high as or even higher than Samsung’s and TSMC’s 7nm technology, but consumers think the latter is more advanced.