Intel Skylake / Kaby Lake

Page 294 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

CakeMonster

Golden Member
Nov 22, 2012
1,497
659
136
So... what about Z270 motherboards if KL release is now delayed? Didn't manufacturers already show them off? Or will there be lower end KL for Z270, or Z270 will just be a motherboard refresh for SL?
 

SAAA

Senior member
May 14, 2014
541
126
116


112W TDP... and no IGP?

My crystal ball says 4.4-4.5GHz base and really limited turbo (think +100-200MHz max), temps might finally be decent enough if it's soldered too.

But aside from Optane support I don't see it as much more interesting than current 6700K, actually it's even less if it doesn't have the IGP and it's new integrated decoders.
It might make 5GHz a thing again but that and perhaps 3-5% IPC over Skylake aren't the end of the world for people with 4.5+GHz Skylakes today.
Definitely an interesting upgrade for anyone still on 2600K or similar tough.
 

Sweepr

Diamond Member
May 12, 2006
5,148
1,143
131
Some unanswered questions:

- There's no Skylake-H GT3e (only Skylake-U GT3e), the Iris Pro versions of Skylake-H are GT4e. Typo?

- 15W Coffee Lake GT3e / 15W Cannon Lake GT2 will coexist. Is Coffee Lake using same gen CPU cores / graphics? Will be interesting to compare 14nm/10nm.

- Any chance the process differences are the reason for this move? Example: if mature/revised 14nm (at least initially) hits higher clocks than 10nm?

- Desktop Coffee Lake socket/chipset compatibility.
 
Last edited:

PaulIntellini

Member
Jun 2, 2015
58
4
71
I guess what I would talk about is Kaby Lake. So one of the things we've learned on 14 nanometers is how to make meaningful performance improvements both in the silicon and then with the silicon combined with the architecture. So we said we already started shipping Kaby Lake to our customers and OEMs. We're seeing meaningful performance across all of the various SKUs of Kaby Lake relative to Skylake. Kaby Lake is built off a Skylake core. And as a result, the die size doesn't significantly grow. So you don't see – there's no driver in the silicon itself to shift the margin structure of this product. We're able to get the performance and feature enhancements with relatively small silicon increases but good improvement on the raw silicon technology itself
BK Q2 2016 earnings call
http://seekingalpha.com/article/399...-results-earnings-call-transcript?part=single
 

mikk

Diamond Member
May 15, 2012
4,245
2,299
136
Means Kabylake will be a bit bigger than Skylake, it's not just a simple refresh.
 

IntelUser2000

Elite Member
Oct 14, 2003
8,686
3,786
136
Interestingly, the leaked Purley slide shows that Xeons would have an optional "Cannonlake Graphics and Media Transcode".

That suggests that either they'll have the Cannonlake graphics block on package, or some SKUs will be offered with Cannonlake graphics block, and some cut.

So we'll potentially get to see 28 core Xeon E5 v5 chips with Iris Pro graphics. More likely we'll see lower end Xeon E5 v5 with Iris Pro, maybe with 14-18 cores.

Seeing that they are aiming Iris Pro GT4e Xeons at video streaming servers, its very plausible that they are bringing such chips to E5 chips as well.
 

Sweepr

Diamond Member
May 12, 2006
5,148
1,143
131
So GameGPU finally added Skylake to their CPU tests, a locked Core i7-6700.

Battlefield 1 - CPU Scaling



http://gamegpu.com/action-/-fps-/-tps/battlefield-1-alpha-test-gpu

Skylake 17.1% faster than Haswell
Haswell 5.7% faster than Sandy Bridge


IBXTGTP = "Intel (R) HD Graphics"
IBXTGTP12 = "Intel (R) HD Graphics"
; KBL
IKBLULTGT1 = "Intel (R) HD Graphics 610"
IKBLULTGT2 = "Intel (R) HD Graphics 620"
IKBLULTGT2F = "Intel (R ) HD Graphics 620 "
IKBLULTGT3 =" Intel (R) HD Graphics 635 "
IKBLULTGT3E15 =" Intel (R) Iris (TM) Graphics 640 "
IKBLULTGT3E28 =" Intel (R) Iris (TM) Graphics 650 "
IKBLULXGT2 =" Intel (R ) HD Graphics 615 "
IKBLDTGT1 =" Intel (R) HD Graphics 610 "
IKBLDTGT2 =" Intel (R) HD Graphics 630 "
IKBLHALOGT1 =" Intel (R) HD Graphics 610 "
IKBLHALOGT2 =" Intel (R) HD Graphics 630 "
IKBLSRVGT2 = "Intel (R) HD Graphics P630"
IKBLWGT2 = "Intel (R) HD Graphics P630"

http://skyline798.blog118.fc2.com/blog-entry-5656.html

Curious if HD Graphics 635 will make it to market, a GT3 iGPU without eDRAM.


Kaby Lake-S / Kaby Lake-H (HD Graphics 630) @ GFXBench

https://gfxbench.com/device.jsp?ben...l&hwtype=iGPU&hwname=Intel(R)+HD+Graphics+630
 
Last edited:

DrMrLordX

Lifer
Apr 27, 2000
22,035
11,620
136
Wow, lots of CPUs beat 60 fps minimum in that bench. And the 5960x walks away with the crown yet again . . . at 3 GHz. Gotta be that l3.
 

EXCellR8

Diamond Member
Sep 1, 2010
3,982
839
136
Would you guys recommend buying the 6700 at this point? Hilarious issues with my AMD system has got me in a tough spot; looking for an upgrade but $350 on a CPU? Ew.

The real irony here is I am posting from my working computer which has a 4790k...
 

sze5003

Lifer
Aug 18, 2012
14,242
638
126
Would you guys recommend buying the 6700 at this point? Hilarious issues with my AMD system has got me in a tough spot; looking for an upgrade but $350 on a CPU? Ew.

The real irony here is I am posting from my working computer which has a 4790k...
I got one coming from an i5 2500k. I had that thing 4+ years and needed to upgrade. I think it's pretty good compared to what i had. Can definitely notice a difference in apps and video conversion.
 

coercitiv

Diamond Member
Jan 24, 2014
6,631
14,069
136
Thanks to clockspeed, sure. Put the 5960x @ 4.0 GHz and it's a different story.
5960x works at 3.3Ghz for MT loads.
6700 works ar 3.7Ghz for MT loads.

getting 5960x at 4Ghz would mean a 21% increase in frequency over stock
getting a 6700k at 4.5Ghz would mean a 21% increase in frequency over 6700 non-k stock.

so, when doing mild overclocking, the status quo will remain mostly similar to what we see at base frequencies (in relative terms ofc).

Skylake might also get some gains from faster DDR4 memory, something the 5960x would not really be sensible to due to increased cache and buffed memory subsystem.

All things being equal, based on the Battlefield 1 Alpha DX11 data, you would need 4.4Ghz+ overclock to pull ahead of Skylake. Things should take a dramatic turn in DX12 though, with 5960x moving from semi-automatic to full machine gun mode
 

YBS1

Golden Member
May 14, 2000
1,945
129
106
5960x works at 3.3Ghz for MT loads.
6700 works ar 3.7Ghz for MT loads.

getting 5960x at 4Ghz would mean a 21% increase in frequency over stock
getting a 6700k at 4.5Ghz would mean a 21% increase in frequency over 6700 non-k stock.

so, when doing mild overclocking, the status quo will remain mostly similar to what we see at base frequencies (in relative terms ofc).

Skylake might also get some gains from faster DDR4 memory, something the 5960x would not really be sensible to due to increased cache and buffed memory subsystem.

All things being equal, based on the Battlefield 1 Alpha DX11 data, you would need 4.4Ghz+ overclock to pull ahead of Skylake. Things should take a dramatic turn in DX12 though, with 5960x moving from semi-automatic to full machine gun mode

You're overlooking the fact that the typical 5960X simply has more overclocking headroom though. The default clocks were set to stay within the desired TDP envelope, not because the 5960X had difficulty going higher. I can bench at 4.8GHz with either my 6700K or my 5960X, the 5960X runs at 4.5 versus the 6700K's 4.6 for safe, sane, daily overclocks. At those speeds the 6700K will have a small per thread advantage, no where near the massive advantage the 5960X will have in anything that can take advantage of 8 cores though.
 

coercitiv

Diamond Member
Jan 24, 2014
6,631
14,069
136
You're overlooking the fact that the typical 5960X simply has more overclocking headroom though.
I'm not overlooking anything, just taking a previous statement made in the thread and taking a look at the numbers. I did not say 5960x cannot go over 4Ghz or 4.4Ghz for that matter, I merely suggested the breaking point at which 5960x pulls ahead in this specific benchmark.

Do you reckon your 5960x will show a massive advantage over your 6700k in BF1 under DX11?
 

Sweepr

Diamond Member
May 12, 2006
5,148
1,143
131
Latest CryEngine game:



Skylake-S slowly distancing itself from Haswell in newer titles.
 
Aug 11, 2008
10,451
642
126
The 6700 is a very strange choice though. Not sure what all core turbo is for the the 6700, but they would get at least a 10% clock boost with a stock K model.

Edit: just saw coercitiv's post. I didnt know the all core turbo was that high. So the K model is only 8% faster. Still though, the 6700 is a strange choice. I expect the best selling models are the 6500 and 6700k. Does anyone know what ram speed they used?
 
Last edited:

nerp

Diamond Member
Dec 31, 2005
9,865
105
106
The 6700 is a very strange choice though. Not sure what all core turbo is for the the 6700, but they would get at least a 10% clock boost with a stock K model.

Edit: just saw coercitiv's post. I didnt know the all core turbo was that high. So the K model is only 8% faster. Still though, the 6700 is a strange choice. I expect the best selling models are the 6500 and 6700k. Does anyone know what ram speed they used?

My non-K 6700 turbos to 4ghz. And it seems to stick there happily under load.
 

coercitiv

Diamond Member
Jan 24, 2014
6,631
14,069
136
My non-K 6700 turbos to 4ghz. And it seems to stick there happily under load.
It probably does so due to the way your BIOS is configured. Default clocks are lower, but even on some non-Z models one may set MT turbo speeds to the same value as 1 core turbo. And if OEMs want to make sure their MB looks good in reviews/benchmarks, they make the setting... default!

I believe Asus calls it "optimized turbo" ()

MultiCore Enhancement: The Debate About Free MHz
The act of telling the processor to run at a certain speed is set by the motherboard, not the processor. So as part of the deal with Intel, motherboard manufacturers’ code in the BIOS the algorithm to make the CPU switch speeds as required. This algorithm can be aggressive, such that turbo boosts are held for a short time when CPU loading goes from low to high, or instant when CPU power is needed or not needed. This algorithm and switching speed can determine how well a motherboard performs in CPU benchmarks.
With these motherboards, usually when XMP is enabled, the CPU is told to use the top turbo boost setting under all loads.

For having a laugh on the subject take a look at this twitter conversation in which David Kanter is trying to find a max stability setting for a Z170 motherboard with help from both Asus and AT senior editor Ian Cutress.
 
Last edited:

YBS1

Golden Member
May 14, 2000
1,945
129
106
I'm not overlooking anything, just taking a previous statement made in the thread and taking a look at the numbers. I did not say 5960x cannot go over 4Ghz or 4.4Ghz for that matter, I merely suggested the breaking point at which 5960x pulls ahead in this specific benchmark.

Do you reckon your 5960x will show a massive advantage over your 6700k in BF1 under DX11?

No, I'd guess they'd be indistinguishable. As I eluded to the per thread difference between them will be small and I doubt the additional cores will impact it that much.
 

SAAA

Senior member
May 14, 2014
541
126
116
Latest CryEngine game:



Skylake-S slowly distancing itself from Haswell in newer titles.

Distancing? That's running away full steam compared to almost any bench app, here it's 21% better at minimum and 15% more average frames with similar clocks!

The 6700 is a very strange choice though. Not sure what all core turbo is for the the 6700, but they would get at least a 10% clock boost with a stock K model.

Edit: just saw coercitiv's post. I didnt know the all core turbo was that high. So the K model is only 8% faster. Still though, the 6700 is a strange choice. I expect the best selling models are the 6500 and 6700k. Does anyone know what ram speed they used?

It's pretty nice to compare generations against each other though, heck Skylake is 42-45% over Sandy i7 here... I don't know what ram they were using but over time ddr4 4GHz is only going to be a lot cheaper and that will make Skylake/Kabylake even faster! :biggrin:
 

DrMrLordX

Lifer
Apr 27, 2000
22,035
11,620
136
5960x works at 3.3Ghz for MT loads.
6700 works ar 3.7Ghz for MT loads.

Interesting point, though just looking at the slides, it's impossible to know exactly what clockspeeds were reached during the benchmark. Did they mean that the displayed clockspeeds were the actual ones reached during the bench?

Or are they just being lazy and posting the base clocks for the CPUs?

Skylake might also get some gains from faster DDR4 memory, something the 5960x would not really be sensible to due to increased cache and buffed memory subsystem.

The 5960x might not gain much from faster memory, but the 4770k might. The advantage the 5960x gains from larger L3 hints that fast system memory could be quote a boon for Haswell in that particular benchmark, if approached properly.

Skylake-S slowly distancing itself from Haswell in newer titles.

Definitely pulling ahead of a bog standard 4770k. I would like to see Broadwell-C in the mix.
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |