Intel Skylake / Kaby Lake

Page 139 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

AtenRa

Lifer
Feb 2, 2009
14,000
3,357
136
I've never understood why people focus on today's top cards when talking about a CPU you'll keep for years. Presumably, GPU's will get faster. When they do, you'll want to have the faster CPU to keep up with them.

And games are becoming more GPU limited as well :rolleyes:
 

witeken

Diamond Member
Dec 25, 2013
3,899
193
106
Good clocks from Skylake-U. Confirms also the conspiracies about 14nm to be hugely wrong.
 

AtenRa

Lifer
Feb 2, 2009
14,000
3,357
136
Could someone explain to me why some people want gaming benches to hit GPU bottleneck and be GPU benches when these benches should be about Skylake CPU performance?

Its very nice to see CPU gaming benchmarks at 720p but that is misleading of the real performance you will see in real gaming, you know, 1080p and above with AA filters and higher IQ settings.

He's argument that Skylake is 20-30% faster than Haswell in real gaming is laughable.
 

ShintaiDK

Lifer
Apr 22, 2012
20,378
145
106
Its very nice to see CPU gaming benchmarks at 720p but that is misleading of the real performance you will see in real gaming, you know, 1080p and above with AA filters and higher IQ settings.

He's argument that Skylake is 20-30% faster than Haswell in real gaming is laughable.

Do you remember Tomb Raider benchmark vs ingame test? 400% difference in performance.

I know your mission in life is to show that weak CPUs are "good enough". But they are not. No matter how hard you keep trying day after day.

And what is this?

getgraphimg.php

getgraphimg.php

getgraphimg.php


And that doesnt even include minimum FPS.
 
Last edited:

Tovarisc

Member
Jun 12, 2015
50
0
0
Its very nice to see CPU gaming benchmarks at 720p but that is misleading of the real performance you will see in real gaming, you know, 1080p and above with AA filters and higher IQ settings.

He's argument that Skylake is 20-30% faster than Haswell in real gaming is laughable.

But I still don't quite see your point. Why I should be interested in GPU performance when reading CPU benchmark? I want to know how CPU affects gaming performance if and when game X hits against CPU instead of GPU. If I want to know how Gigabyte 980Ti G1 performs in games then I would read GPU benchmarks.

Skylake maybe doesn't give you 30% FPS increase in "real gaming situations" because most of the time games are GPU limited, but knowing where Skylake stands if game for some reason hits CPU bottleneck is quite nice information to have. When hitting that CPU bottleneck Skylake seems to give very nice numbers.
 

AtenRa

Lifer
Feb 2, 2009
14,000
3,357
136
Do you remember Tomb Raider benchmark vs ingame test? 400% difference in performance.

I know your mission in life is to show that weak CPUs are "good enough". But they are not.

And whats this?

getgraphimg.php

getgraphimg.php

getgraphimg.php


And that doesnt even include minimum FPS.


read again,

http://www.hardware.fr/articles/940-1/intel-core-i7-6700k-i5-6600k-z170-skylake-test.html

Crysis 3 settings : 1920*1080 Very High, sans anti-aliasing.
ARMA III settings : 1920*1080 Ultra, sans anti-aliasing.
x-plane 10 settings : 1920*1080 sans anti-aliasing.
F1 2013 settings : 1920*1080 sans anti-aliasing.
Watch Dogs settings : 1920 * 1080 with a level of overall quality Ultra without anti-aliasing.
Total War Rome II Settings : 1920 * 1080 Extreme but by disabling the AA and SSAO.
Company Of Heroes II Settings : 1920 * 1080 maximum quality made exception of anti-aliasing.
Anno 2070 settings : 1920 * 1080 with very high settings and still no anti-aliasing.

That is not how you going to play with your GTX980, the difference from 4790K vs 6700K at default is negligible.
 

ShintaiDK

Lifer
Apr 22, 2012
20,378
145
106
You forgot to show the higher resolutions if you wanted to have an actual point. But I think we all know why you didnt.
 

AtenRa

Lifer
Feb 2, 2009
14,000
3,357
136
But I still don't quite see your point. Why I should be interested in GPU performance when reading CPU benchmark? I want to know how CPU affects gaming performance if and when game X hits against CPU instead of GPU. If I want to know how Gigabyte 980Ti G1 performs in games then I would read GPU benchmarks.

Skylake maybe doesn't give you 30% FPS increase in "real gaming situations" because most of the time games are GPU limited, but knowing where Skylake stands if game for some reason hits CPU bottleneck is quite nice information to have. When hitting that CPU bottleneck Skylake seems to give very nice numbers.

You want to know how the CPU will perform in the settings you are going to play the game, that is at 1080p(or higher) with high IQ settings and AA filters if your GPU allows it. Not at 720p with low settings.

And from various reviews it seems that with current games and GPUs there is almost zero difference, in real gaming scenarios, between Haswell and Skylake.
 
Last edited:

Sweepr

Diamond Member
May 12, 2006
5,148
1,142
131
Could someone explain to me why some people want gaming benches to hit GPU bottleneck and be GPU benches when these benches should be about Skylake CPU performance?

Remmember some people are on a mission to prove that CPUs with absolute crap per core performance like AMD's FX lineup can keep up with best in class x86 performance (Skylake). This misleading idea is only possible if you handpick prescripted benchmarks, preferably with crancked up resolution or filters to force a GPU bottleneck all the time. Thank god some websites are still testing how brand new CPU would fare if you actually needed the extra performance provided by them (which is the main reason why you replace a CPU in the first place).


Right now there's a huge gap between Intel and AMD in terms of performance per clock. PCLab included results for both Core i7 6700K and FX8350 operating at 4.7GHz in their review so it's easy to make an architecture comparison. According to their results, Skylake-S is:

34,3% faster @ Assassins Creed Unity
60,4% faster @ Arma III
60,3% faster @ Battlefield 4 Multiplayer
64% faster @ Counter Strike Global Offensive
32.5% faster @ Crysis 3
104% faster @ Far Cry 4
114% faster @ GTA V
85% faster @ The Witcher 3
60.6% fater @ Watch Dogs
87.4% faster @ Project Cars
133% faster @ Starcraft 2
139.1% faster @ Total War Attila

Overall: Skylake-S 81.2% faster than Vishera per clock.

gry.png


Untitled-1_zps68rzxsyb.jpg


http://forums.anandtech.com/showthread.php?t=2428363&page=131

Eurogamer said:
But in the here and now, the Core i5 6600K is a strong product - at worst it provides mildly enhanced performance over its predecessors, at best it's noticeably faster and should help to reduce CPU bottlenecking during gameplay. For those looking to buy or construct a new, capable gaming PC, the i5 remains the default choice and the 6600K is the best iteration yet - AMD's older eight-core FX chips are cheaper, but our tests clearly demonstrate that despite the low-power many-core set-up of the new console CPUs, modern games generally prefer the high per-core performance that Intel provides. Impressive stuff overall then, which begs the question, just how good is the Skylake 6700K? Let's just say that the initial data already looks very impressive.

gta5.png


fc4.png


tw.png


Eurogamer said:
Our benchmarks above don't fully replicate the gameplay experience - they are designed to compare CPU architectures with graphics reduced as a limiting factor, but the bottom line is that a growing number of titles are seeing the CPU act as the bottleneck - a fully maxed GTA 5 is CPU-limited even on an overclocked Core i7 4790K, while certain scenes in the Witcher 3 on ultra settings can also cause difficulties for even the most powerful processor. Our contention is that more games like this will come along, and in demanding CPU titles, the Core i5 6600K will make a noticeable difference.

Edit: Nevermind the fact that there's more to game testing than average framerates:

pcars-99th.gif


pcars-8ms.gif


civ-16ms.gif


fc4-16ms.gif


Here's were crap gaming CPUs literally fall flat.
 
Last edited:

dahorns

Senior member
Sep 13, 2013
550
83
91
And games are becoming more GPU limited as well :rolleyes:

I don't know what that means. Presumably you could at some point upgrade your GPU, perhaps even to the point of no longer being GPU limited, right? This seems especially true given the platform improvements offered by Skylake.

Further, we know that games can become CPU limited in actual game environments (e.g., large multi-player situations) rather than staged benchmarks
 

AtenRa

Lifer
Feb 2, 2009
14,000
3,357
136
I don't know what that means. Presumably you could at some point upgrade your GPU, perhaps even to the point of no longer being GPU limited, right? This seems especially true given the platform improvements offered by Skylake.

Further, we know that games can become CPU limited in actual game environments (e.g., large multi-player situations) rather than staged benchmarks

We are not talking about a Pentium G3258 vs Core i7 6700K here, we are talking about Haswell Core i7 4790K vs Skylake Core i7 6700K. His argument was that Skylake is 20-30% faster in gaming than same SKU Hasswell, witch is not true in real gaming scenarios.

He will gain much more going for a GPU upgrade from his GTX980 to GTX980Ti than upgrading the entire platform(Mobo + Memory + CPU). For less than a platform upgrade, he could buy a Core i7 4790K + a GTX980Ti and by selling his older Hardware (4670 + GTX980) the upgrade would be very $ efficient.
 

CHADBOGA

Platinum Member
Mar 31, 2009
2,135
832
136
But I still don't quite see your point. Why I should be interested in GPU performance when reading CPU benchmark? I want to know how CPU affects gaming performance if and when game X hits against CPU instead of GPU. If I want to know how Gigabyte 980Ti G1 performs in games then I would read GPU benchmarks.

Skylake maybe doesn't give you 30% FPS increase in "real gaming situations" because most of the time games are GPU limited, but knowing where Skylake stands if game for some reason hits CPU bottleneck is quite nice information to have. When hitting that CPU bottleneck Skylake seems to give very nice numbers.

Look at it from AtenRa's perspective.

If you get to see Skylake's(or Haswell's) performance when it isn't being held back by the GPU, you would see how poorly AMD processors perform by comparison and how bad they will be in the future.
 
Last edited:

CHADBOGA

Platinum Member
Mar 31, 2009
2,135
832
136
Let me guess Kaby Lake will be 5% faster than Skylake and Cannonlake will be another 5% faster than Kaby Lake. We won't be getting any exciting desktop CPU launch from Intel for a long time.
AMD launches on the other hand are going to be a fun ride with Carrizo and Zen coming up.
LOL Logic FAIL

@ShintaiDK
you wasted your money because there is going to be negligible difference in your gaming experience going from Haswell i5 to Skylake i7 but it's your money so you can waste it as much as you please.

He is replacing an Ivy Bridge machine for a Skylake.
 

dahorns

Senior member
Sep 13, 2013
550
83
91
We are not talking about a Pentium G3258 vs Core i7 6700K here, we are talking about Haswell Core i7 4790K vs Skylake Core i7 6700K. His argument was that Skylake is 20-30% faster in gaming than same SKU Hasswell, witch is not true in real gaming scenarios.

We are?

No.

But again you forgot to do your homework.

4670 vs 6700K.

He will gain much more going for a GPU upgrade from his GTX980 to GTX980Ti than upgrading the entire platform(Mobo + Memory + CPU). For less than a platform upgrade, he could buy a Core i7 4790K + a GTX980Ti and by selling his older Hardware (4670 + GTX980) the upgrade would be very $ efficient.

I assume he could both buy the 6700k and a new gpu? Or maybe he could upgrade the gpu when the next line of kickass GPUs arrives? Why are you assuming people building high end systems are looking for the absolute best performance per dollar?
 
Last edited:

AtenRa

Lifer
Feb 2, 2009
14,000
3,357
136

Yes,

Same as 4670 vs 4790K :whiste:

as for,

I assume he could both buy the 6700k and a new gpu? Or maybe he could upgrade the gpu when the next line of kickass GPUs arrives? Why are you assuming people building high end systems are looking for the absolute best performance per dollar?

I dont have a problem with people want to upgrade to Skylake for the platform or if they are upgrading from Sandy etc, my debate is about his claim that Skylake is 20-30% faster in Gaming against Haswell, witch is not true.
 

dahorns

Senior member
Sep 13, 2013
550
83
91

????? Obviously you are talking about it. I don't see anyone else doing it

I dont have a problem with people want to upgrade to Skylake for the platform or if they are upgrading from Sandy etc, my debate is about his claim that Skylake is 20-30% faster in Gaming against Haswell, witch is not true.

Where did he claim that? He said that the 6770k would offer 20-30% more performance than his 4670. That seems probably true? You asserted that the percentage would be the same for the 4790k. He then provided charts showing that the 6700k offers superior performance to the 4790k (though not 20-30% superior).

Basically, it seems you are having a conversation with yourself, which is a bit confusing for the rest of us.
 

phillyman36

Golden Member
Jun 28, 2004
1,762
160
106
ShopBLT is showing 9/03 for 6700k. I hope it shows up somewhere else like Newegg or Amazon because I got my motherboard and ram on 8/5 and the wait is killing.

http://www.shopblt.com/cgi-bin/shop...110030005034_B1G7022P.shtml&order_id=!ORDERID!

I feel your pain my brother lol. The only thing with shopblt is there is no notify option. You have to pre order or else you are at the end of the line. I was looking at a store in the uk and did the online dollar conversion. Its came to $548.00. No thanks

Also found this in stock
http://www.memory4less.com/m4l_itemdetail.aspx?itemid=1478953688&partno=i7-6700K

$913.73 Way to try to take advantage of someone. If they sell any hope they feel good about themselves.
 

cmdrdredd

Lifer
Dec 12, 2001
27,052
357
126
I got i5 750 @ 3,8GHz way back when and it has kept me in gaming all the way to this day. Either ancient Lynnfield is somehow super good CPU or i5 has more than enough juice for gaming. Now it's time to upgrade to i5 6600K tho.


There have been times when I thought the extra threads would have been useful that's all. I should have went with the hyperthreading chip.
 

BSim500

Golden Member
Jun 5, 2013
1,480
216
106
Could someone explain to me why some people want gaming benches to hit GPU bottleneck and be GPU benches when these benches should be about Skylake CPU performance?

Remember some people are on a mission to prove that CPUs with absolute crap per core performance like AMD's FX lineup can keep up with best in class x86 performance (Skylake). This misleading idea is only possible if you handpick prescripted benchmarks, preferably with crancked up resolution or filters to force a GPU bottleneck all the time. Thank god some websites are still testing how brand new CPU would fare if you actually needed the extra performance provided by them (which is the main reason why you replace a CPU in the first place).
^ The answer is this, Tovarisc. Recurring brand fanboyism plays a large part. It's why you see certain regulars state 'That is not how the average gamer is going to play' one minute for 1080p + Ultra + 0x AA benchmarks for the settings being "too low" on Intel chips, and then do AMD APU benchmarks declaring A8-7600 at 20-30fps @ 720p + "Low" presets + the same 0AA as 'acceptable performance for the average gamer who doesn't own a GPU' the next. "The average gamer's" settings, resolution & other non-CPU components can change like the wind in these forums when it's Intel vs AMD CPU's being tested... :D

You are correct - for CPU focused benchmarks, yes you generally should try and eliminate GPU bottlenecks precisely because these days, people tend to upgrade their GPU more often vs their CPU, so any current GPU bottlenecks are likely to be removed 1-2 years from now, and benchmarking without GPU-bottlenecking today more clearly reveals the predicted endurance / overhead of the CPU over a 4-5 year potential lifespan towards the 2nd half of the CPU's life. If people want to do "real world gameplay" (ie, GPU crippled "CPU" benchmarks), to see if the upgrade is worth it with a slower GPU, by all means do them in addition to proper CPU benchmarks (no GPU bottleneck), not instead of.

"Proper" game settings are also entirely subjective. If someone wants to use SMAA or FXAA (only 5-10% GPU impact instead of 30-60% of 4-8x MSAA), to get higher fps all round / personal preference, they are not wrong to do so as long as they benchmark the same settings for all chips in the test. Same goes for disabling annoying Ultra preset default on cr*p like Chromatic Abhorration, Vignetting, etc, (which many of us regularly do regardless of performance / hardware brand). One man's "It's cheating to turn that off because it lowers GPU usage which makes faster CPU's with more overhead look even better" is another man's "well I turn it off in real life and I think it's even more cheating to artificially overload a GPU with cr*p I won't use in practice to 'accidentally on purpose' cap the fastest CPU's during CPU game benchmarking to make the slower ones look relatively better". ;)
 
Last edited: