On Aug. 21, Intel (INTC +0.64%) is expected to announce its eighth-generation Core processor family for the personal computer market. Most of these components are expected to be manufactured using the company's third version of its 14nm technology, branded 14nm++.
The performance and power efficiency of a chip-manufacturing technology is a key driver of the performance and power efficiency of a processor. Intel said at its Technology and Manufacturing Day earlier this year that its 14nm++ technology can deliver up to 26% better performance at the same power or the same performance at 52% less power than the first iteration of its 14nm technology, introduced back in late 2014 .
Image source: Intel.
Intel's ninth-generation Core processor family, which could arrive in the second half of 2018 or in the first half of 2019 (Intel hasn't said), will be called Ice Lake, and has long been expected to be manufactured using some flavor of the company's upcoming 10nm technology.
Thanks to (a potentially inadvertent) posting on Intel's website spotted by user "nvgpu" on the AnandTech forums, we now know precisely which flavor of 10nm technology the Ice Lake chips will be built using.
It's 10nm+
"The Ice Lake processor family is a successor to the 8th generation Intel Core processor family," Intel says on its website. "These processors utilize Intel's industry-leading 10 nm+ process technology."
Recall that at Intel's Technology and Manufacturing Day, the company said that it's planning three generations of its 10nm technology -- 10nm, 10nm+, and 10nm++.
The initial version of 10nm, which will be used to manufacture the company's Cannon Lake processor family for 2-in-1 notebooks, is expected to deliver slightly better performance than the 14nm+ technology used to build the company's seventh-generation Core processors.
Per Intel, 10nm+ should offer transistor performance that's almost on par with the 14nm++ technology that will be used to build most of this year's eighth-generation Core chips, though with the benefit of lower capacitance (Intel says that this translates into lower power consumption) as well as dramatically reduced transistor area (which means that more transistors -- that is, the individual elements that computer chips are made of -- can be crammed into a given area, increasing features and functionality).
What this means for Ice Lake
Although the 10nm+ technology that Ice Lake will be manufactured on may regress slightly from 14nm++ in terms of transistor performance (this would potentially translate into slightly lower peak chip frequency capability), Intel should be able to dramatically improve the performance and features of Ice Lake relative to the prior generation parts through other means.
The smaller transistors should allow Intel to make substantial increases in the sheer number of transistors that it throws are the various sub-systems within the Ice Lake chips, improving performance significantly while keeping chip size (cost) in check.
Intel is also likely to have found "smarter" ways to do things in its chips (that's why Intel pays so many sharp engineers), translating into an improved architecture with Ice Lake that'll lead to better performance and features irrespective of the underlying manufacturing technology used.
