Slashdot Mirror
Intel Officially Reveals Post-8th Generation Core Architecture Code Name: Ice Lake, Built On 10nm+ (anandtech.com)
Intel has confirmed the existence of a new processor family called Ice Lake that will be made on Intel's 10nm+ process. The company published basic information on the Ice Lake architecture on their codename decoder. AnandTech reports: This is an unexpected development as the company has yet to formally detail (let alone launch) the first 10nm Core architecture -- Cannon Lake -- and it's rare these days for Intel to talk more than a generation ahead in CPU architectures. Equally as interesting is the fact that Intel is calling Ice Lake the successor to their upcoming 8th generation Coffee Lake processors, which codename bingo aside, throws some confusion on where the 14nm Coffee Lake and 10nm Cannon Lake will eventually stand. As a refresher, the last few generations of Core have been Sandy Bridge, Ivy Bridge, Broadwell, Haswell, Skylake, with Kaby Lake being the latest and was recently released at the top of the year. Kaby Lake is Intel's third Core product produced using a 14nm lithography process, specifically the second-generation '14 PLUS' (or 14+) version of Intel's 14nm process.
Working purely on lithographic nomenclature, Intel has three processes on 14nm: 14, 14+, and 14++. As shown to everyone at Intel's Technology Manufacturing Day a couple of months ago, these will be followed by a trio of 10nm processes: 10nm, 10nm+ (10+), and 10++. On the desktop, Core processors will go from 14 to 14+ to 14++, such that we move from Skylake to Kaby Lake to Coffee Lake. On the Laptop side, this goes from 14 to 14+ to 14++/10, such that we move from Skylake to Kaby Lake to Coffee Lake like the desktops, but also that at some time during the Coffee Lake generation, Cannon Lake will also be launched for laptops. The next node for both after this is 10+, which will be helmed by the Ice Lake architecture.
Working purely on lithographic nomenclature, Intel has three processes on 14nm: 14, 14+, and 14++. As shown to everyone at Intel's Technology Manufacturing Day a couple of months ago, these will be followed by a trio of 10nm processes: 10nm, 10nm+ (10+), and 10++. On the desktop, Core processors will go from 14 to 14+ to 14++, such that we move from Skylake to Kaby Lake to Coffee Lake. On the Laptop side, this goes from 14 to 14+ to 14++/10, such that we move from Skylake to Kaby Lake to Coffee Lake like the desktops, but also that at some time during the Coffee Lake generation, Cannon Lake will also be launched for laptops. The next node for both after this is 10+, which will be helmed by the Ice Lake architecture.
Intel has two "problem" architectures - Broadwell and Cannonlake. Both suffer from debuting on a new and untested process. Broadwell took forever and a day to make it out of the mobile sector, eventually finding its way into Xeon-D, Xeon, and HEDT products as well (not counting the mostly-unsupported and barely-mentioned Broadwell-C and Broadwell-R processors). Skylake and Kabylake were basically on the same process with uarch tweaks. Okay, Kabylake is on 14nm+ but whatever.
If you will recall, delays in the 14nm process caused Intel to re-release Haswell as Devil's Canyon, otherwise known as the i7-4790k and i5-4690k processors or "Haswell Refresh".
Intel is experiencing delays with 10nm and Cannonlake which is meant to be its debut architecture. They are using Coffeelake as a "Kabylake Refresh" since it uses the same core architecture as Kaby (and Skylake; in fact, both Kabylake and Coffeelake are Skylake refreshes). It uses a respun version of 14nm+ - 14nm++ or whatever you want to call it - and it features up to two more cores than what you can get with Kabylake.
So Cannonlake is only a definite go for mobile whenever Intel is finally ready to start selling 10nm CPUs. That could be a bit.
We may eventually see Xeon-D products, Xeon products, and HEDT products based on Cannonlake. On the desktop, we should not expect Cannonlake at all, not even a successor to the controversial Broadwell-C. So just as most Intel buyers made the jump from Haswell to Skylake, now we're going to make the jump from Skylake/Kabylake/Coffeelake to Icelake. They skipped Broadwell on the desktop (mostly), and now they're skipping Cannonlake.
The other thing that's really confusing is that Intel hasn't actually released a new uarch in awhile. Their last "new" uarch was Skylake. Skylake-X - the HEDT/server version of Skylake - is nothing but Skylake with bolt-on AVX512 functionality and a rejiggered cache configuration to make it more competitive in certain server workloads (as a consequence, Skylake-X is slower in games at any given clockspeed than Skylake/Kabylake at the same clock). All signs point to Coffeelake being planned all along as a die-shrink of Skylake to 10nm. So Skylake, Kabylake, Coffeelake, and (probably) Coffeelake ARE ALL ACTUALLY THE SAME UARCH. Weird huh?
Icelake will be the first "new" uarch Intel will have released since the debut of Skylake.
Ryzen 1700. It's about the same price as the 2600K was when it launched but you get twice as many cores and threads. It won't overclock as well as a 2600K, but the performance per clock is going to be slightly better in most cases. Really though, a 2600K is still going to be a fine CPU and unless you have any real need to upgrade you can stick with it. The 2600K also has onboard graphics that the Ryzen CPU lacks if that's a deal breakerl.
Intel sat on their hands for multiple generations while AMD had nothing to offer and now they're getting bit in the ass now that they have to compete again. They really should have made 6-core mainstream parts several generations ago.
This mentions 14+ and 14++ - https://en.wikichip.org/wiki/14_nm_lithography_process
Pluses means 'slightly smaller' and/or more mature/stable, generally allowing greater chip yield and higher clock speed.
Corruption is convincing someone that the selfless ideal is the same as their selfish ideal.
It's even funnier that they're trying to focus on things past their own next-future product which isn't even out either. Can you imagine Toyota telling us about their new 2019 cars in mid-2017?
Well the devil's advocate could say that you want them to only focus on next quarter's profits. Intel has always talked to investors and technology analysts about their road map, their code names and process architecture has never been directed at consumers. Let's compare them to the competition, how long before the release of an actual Zen processor did AMD start to hype it up? Announced in May 2015, released in March 2017. Or if you want to do the car analogy, how long from Musk announced the Model 3 until the first one rolled off the assembly line? July 2014 to July 2017.
The reason we are discussing Intel's road map now is that we don't quite believe them anymore. They used to have a tick-tock, now it's like 2014 14nm, 2015 14nm, 2016 14nm+, 2017 14nm++. Those two years that were supposed to become three (process-architecture-enhancement) is now four, sure it's damage control. But it's not really damage control directed at consumers, this is talking to the stock market saying we've been enhancing the product we have and we got a plan. Nothing particularly unexpected here, but exclude the 10nm and redraw the progress lines and they're slipping...
Live today, because you never know what tomorrow brings