With the introduction of Intel's 11th generation Rocket Lake processors, the company's long and painful run at 14 nm has finally come to an end. Intel has already announced that Rocket Lake will be the last desktop processor to use the 14nm node, which will be followed by the 10nm Alder Lake chips later this year.
Based on the reviews already, Rocket Lake 14nm will end up with a bit of a whimper. The gains over previous iterations aren't as high as expected, and the reduction to just eight cores on the high end didn't help. With the speed at which its successor will chase, Intel likely knows it has to keep going to survive. Fast.
With 14 nm in the dust and a fresh supply of technical investments ahead of us, Intel is finally completing its seven-year transition to 10 nm.
From tick-tock to tick-tock
Intel released products in a tick-tock pattern that was first introduced in 2007. This meant that Intel would downsize its chip size every two years. Smaller transistors mean more transistors – all with the aim of increasing efficiency, price and performance. That fits in well with the pace of innovation at Moore & # 39; s Law and the last twenty years of processor development.
However, that changed in 2016. Cannon Lake was supposed to be Intel's first 10nm chip, originally planned for 2016. Instead, this year the company introduced its Kaby Lake processors. Instead of switching from 14 to 10 nm as a “token”, the company had started repeating or “updating” its 14 nm node year after year. This is how the delays in the changeover to 10 nm began, initially from 2015 to 2017. In view of Intel's excellent positioning compared to its competitors, nobody turned a blind eye.
But then Cannon Lake was postponed for another year to 2018. And when it finally started we found out how bad the situation really was.
Cannon Lake, the first 10nm processor to start in just one configuration: the Core i3-8121U. This version for ultra-small volume laptops was a preview of how long the full transition to 10nm would take. It's not exactly the confident move we've been waiting three years for. To support the demand for an actually updated launch of laptop processors, Intel had to release its 8th generation Whiskey Lake processors instead.
It would be two more years before a 10nm successor to Cannon Lake known as Ice Lake would hit the market. It was a big moment for Intel – real 10nm processors in real high-end laptops that people could buy. It came with a new (and even more confusing) naming scheme, a renewed emphasis on improved built-in graphics, and some modest performance gains over the 14nm parts.
However, there were two problems. First, the clock speeds were very low and the volume was still missing. Intel had to release another 14nm counterpart (codenamed Comet Lake) to meet market demand. In addition, the low frequencies limited the release to thin and light laptops only. Anything over 28 watts, such as B. gaming laptops or desktops, stayed at 14 nm.
It's finally over. Or is it?
Intel is still in this situation today. Intel has slowly ramped up production to 10 nm so that the low wattage laptop chips can be completely switched from 14 nm to 14 nm. The new 11th generation Tiger Lake is 10nm laptops, and most of the Intel laptops you can buy in 2021 will have a 10nm chip.
And 45-watt Tiger Lake-H processors are slated to come soon, which will complete the laptop journey down to 10 nm. Meanwhile, Alder Lake 12th generation chips will end the search for 10 nm on the desktop side of history later this year.
But as with any technology, companies must never sit on their hands, and 10nm is just one stop on the road where Intel's massive $ 20 billion investment comes into play.
The company seems to know that it cannot afford another such delay. Intel has pushed back towards a tick-tock release plan, announcing plans to bring 7nm production in 2023.
That doesn't mean that Intel is suddenly back on top. AMD and Apple have a head start, and it's a performance gap that will continue to be an issue for Intel. But for the first time in recent years, Intel is back on track – and the death of 14 nm bodes well for the future.