Monday, October 14th 2019
Intel Scraps 10nm for Desktop, Brazen it Out with 14nm Skylake Till 2022?
In a shocking piece of news, Intel has reportedly scrapped plans to launch its 10 nm "Ice Lake" microarchitecture on the client desktop platform. The company will confine its 10 nm microarchitectures, "Ice Lake" and "Tiger Lake" to only the mobile platform, while the desktop platform will see derivatives of "Skylake" hold Intel's fort under the year 2022! Intel gambles that with HyperThreading enabled across the board and increased clock-speeds, it can restore competitiveness with AMD's 7 nm "Zen 2" Ryzen processors with its "Comet Lake" silicon that offers core-counts of up to 10.
"Comet Lake" will be succeeded in 2021 by the 14 nm "Rocket Lake" silicon, which somehow combines a Gen12 iGPU with "Skylake" derived CPU cores, and possibly increased core-counts and clock speeds over "Comet Lake." It's only 2022 that Intel will ship out a truly new microarchitecture on the desktop platform, with "Meteor Lake." This chip will be built on Intel's swanky 7 nm EUV silicon fabrication node, and possibly integrate CPU cores more advanced than even "Willow Cove," possibly "Golden Cove."The HardwareLuxx article making these explosive revelations attributes the sudden change in Intel's plans to the company not being able to scale clock-speeds of "Ice Lake" high enough to establish product leadership. It feels "Skylake," which has IPC parity with "Zen 2," has enough scalability and clock-speed headroom to stay competitive with AMD at high clock-speeds. The company will augment next-generation uncore (revamped memory controllers, support for PCIe gen 4.0, Gen12 iGPU, etc.), with "Skylake" CPU cores, over time. Other areas where Intel could grow its mainstream desktop silicon is cache rebalancing similar to its HEDT chips, and implementing the Mesh Interconnect to maintain low latencies as core-counts enter two-figures.
Interestingly, 10 nm "Ice Lake" remains on Intel's enterprise roadmap, where the company appears more desperate not to cede market-share to AMD, especially as businesses around the world set their 5G plans rolling, springing a cycle of hardware updates in the data-center. 2020 could see the introduction of Xeon Scalable processors based on 10 nm "Ice Lake" microarchitecture with "Sunny Cove" CPU cores. In 2021, the company will introduce the "Sapphire Rapids" Xeon processor with even more cores and larger I/O connectivity, spearheaded with PCI-Express gen 5.0.
Update Oct 15th: Intel has released a statement, denying these claims, read more here.
Source:
HardwareLuxx.de
"Comet Lake" will be succeeded in 2021 by the 14 nm "Rocket Lake" silicon, which somehow combines a Gen12 iGPU with "Skylake" derived CPU cores, and possibly increased core-counts and clock speeds over "Comet Lake." It's only 2022 that Intel will ship out a truly new microarchitecture on the desktop platform, with "Meteor Lake." This chip will be built on Intel's swanky 7 nm EUV silicon fabrication node, and possibly integrate CPU cores more advanced than even "Willow Cove," possibly "Golden Cove."The HardwareLuxx article making these explosive revelations attributes the sudden change in Intel's plans to the company not being able to scale clock-speeds of "Ice Lake" high enough to establish product leadership. It feels "Skylake," which has IPC parity with "Zen 2," has enough scalability and clock-speed headroom to stay competitive with AMD at high clock-speeds. The company will augment next-generation uncore (revamped memory controllers, support for PCIe gen 4.0, Gen12 iGPU, etc.), with "Skylake" CPU cores, over time. Other areas where Intel could grow its mainstream desktop silicon is cache rebalancing similar to its HEDT chips, and implementing the Mesh Interconnect to maintain low latencies as core-counts enter two-figures.
Interestingly, 10 nm "Ice Lake" remains on Intel's enterprise roadmap, where the company appears more desperate not to cede market-share to AMD, especially as businesses around the world set their 5G plans rolling, springing a cycle of hardware updates in the data-center. 2020 could see the introduction of Xeon Scalable processors based on 10 nm "Ice Lake" microarchitecture with "Sunny Cove" CPU cores. In 2021, the company will introduce the "Sapphire Rapids" Xeon processor with even more cores and larger I/O connectivity, spearheaded with PCI-Express gen 5.0.
Update Oct 15th: Intel has released a statement, denying these claims, read more here.
148 Comments on Intel Scraps 10nm for Desktop, Brazen it Out with 14nm Skylake Till 2022?
Incentives is OK to a certain extent, but rewards for knocking out the competition is illegal. No fix will mitigate Intel's CPU malware and security issues. It's was a deliberate design technique to squeeze out more performance and years later got caught.
Intel needs a new design built from the ground up. That's most likely why they hired Jim Keller. I can see something new by 2023 to 2025.
That's the fundamental background of this Intel-AMD battle.
AMD can make very advanced designs that depend on very limited technology - because they're fine selling 10% of what Intel does.
Intel has to design CPUs that they can make in quantities their partners demand.
That's why the existing 10nm supply goes to fairly low-volume products: mobile multimedia SoCs and Nervana.
Intel could redesign Skylake for TSMC's 7nm, but what's the point? I doubt it'd even saturate demand coming from Dell alone. Speculative execution was invented in the 80's and utilized by almost all CPU makers - including AMD that you worship so much.
And yes: it was expected that this technique will make some type of attack possible, but no one managed to exploit it until 2016.
It happened, se we had to sacrifice some of the gains. That's all. Stop whining. :)
Nobody is whining but those that prefer Intel CPU's and Intel themselves.
You like to talk about speculative execution. Why not spend an evening learning what it is?
That would be nice and yield the 10% gain without changing anything except frequency.....
www.phoronix.com/scan.php?page=article&item=3900x-9900k-mitigations&num=1
9900K means Meltdown is taken care of in hardware, so it's Spectre vs Spectre.
Phoronix also did a few comparisons of earlier CPUs (also Xeon vs EPYC), where Intel side has both Spectre and Meltdown mitigations (AMD just Spectre).
Of course, we know next to nothing about Intel's 7nm node at this time... who's to say they won't have the same problems with their 7nm node?
Don't forget the other ones other than spectre and meltdown: Foreshadow, ZombieLoad, Fallout, RidL
Also remember that most of these issues were known when they released their 7th and 8th Gen Processors... So they released a knowingly broken product.
/deadpan
get your facts straight pls!!
Mainstream desktops (4-6 cores) can remain on 14nm.
High-end desktop is a weird market. With current margins it isn't very attractive anymore and Intel would love to dump it. If AMD is fine with those margins - let them have it.
But that would only work in an ideal world where people do conscious decisions based on full information. And the world it's not like that.
In reality, high-end desktop is doing most of the PR. These CPUs are covered in reviews and that's what people read and talk about.
Also, everyone knows a PC geek and he will likely have a powerful desktop - and that's the person we turn to for PC advice.
So Intel will try to stay relevant in this game by all means possible - most likely pushing HEDT down to make it more popular. And lowering prices.
Because $1 lost on desktops may mean $3 earned somewhere else. Even if their 10nm can't do high frequencies, it's not a big deal for their product profile. Mobile and servers will work fine.
It was a bit different with AMD and Zen/Zen+. AMD focuses on gamers. They were very dependent on TSMC providing chips with higher clocks. But most desktops are OEM home/business machines. Intel can cover that with 14nm.
You're really thinking about high-end gaming desktops (the stuff forums like this one focus on). And that market is relatively small for Intel. Because they started very early with 10nm. They were alone and they had some ideas that didn't work well. But it's a big investment, so they kept working on it...
With Intel there's always the issue of scale. They don't need a 10nm node that just works in some products. They need a 10nm node that will work for 400 mln - very diversified - chips yearly.
In the meantime TSMC and Samsung caught up and went straight past.
Intel's 7nm is expected to be launched with competing 5nm for Samsung and TSMC. No rushing this time. Let's hope it works. :) I'm pretty sure there are other tech companies in the world that moved from dog shit level to actually making some money.
The fact that a huge, mainstream company like AMD moved from $2 in 2015 to $30 in 2019 is not something to be proud of.
Price and core counts sure are important, but they are far from everything that matters.
Multiple sources state that ZEN 3 could see IPC gains of over 8% and up to 200 MHz faster clock speeds over ZEN2.
Looking forward to ZEN3, as it stands that may be my next upgrade. At the moment, my ZEN is doing just fine in and out of PC Gaming.... :)
Never do I use all 16 threads.
Never does it get hot unless I bench on it or F@H ect.
And lately, been running a low pwer state which enables me to passively cool the processor.
Still games even at 3.0ghz 0.9v.
Still maintains max settings CODBO IV and around 80 fps at 1080p.
Good stuff.
But die size? Core count? IPC?
It's really sad that people care about such things. It shows that the group that likes to mock marketing is just as susceptible as everyone else. Why would you be 100% happy if you can't use all threads? :o
It more than covers my needs.
Very good effeceincy in lower power states.
And thats all I can really ask for in an AMD product.
Yep 100% satisfied.
And my 1700x running stock speed is more than enough for my needs.
Runs cool and quiet. .
Ryzen's would benefit a lot more if they somehow prioritized latency.
Agreed.
But we know for sure that Intel is buying a lot of 7nm EUV equipment from ASML. So something will be launched. :) That's not a huge achievement when you think that Ryzen was the first good AMD design since 2007's K10.
And 6 years since the awful Bulldozer.
If you compare Intel CPUs from 2017-2018 and 2011 (not to mention 2007), the architecture is similar, but the performance and efficiency are in 2 different worlds as well. No one says it's a bad CPU. But you paid for 8C/16T and you said yourself you never used 16 threads (and you said it like if it was good...?)
Would you be so happy about 2700X benchmark results (Cinebench etc) if they were 10-20% lower? :)
So the question I asked was: wasn't this a suboptimal choice? Maybe 2600X would be better for you?