Intel to launch 6 core Coffee Lake-S CPUs & Z370 chipset 5 October 2017

gavinh87 · 21 Sep 2017 at 18:19

voidshatter said:
I read from somewhere that the leak was not using the correct version of Cinebench. That's why I'd prefer to see the final numbers.

He's also using a VERY old version of cpuz. 1.74
Not that it has any affect on performance but shows they are using outdated software.

Doobedoo · 21 Sep 2017 at 18:22

gavinh87 said:
I have no clue but he has provided screens and what does he have to lose? If 8pack broke NDA it would have consequences for him and OCUK, some random guy on a forum doesn't have that worry.

Well I'm sceptical as I'm sure you can understand. He provided a screenshot of 6 cores running at 5.2ghz. That is it!

Doobedoo · 21 Sep 2017 at 18:23

gavinh87 said:
He's also using a VERY old version of cpuz. 1.74
Not that it has any affect on performance but shows they are using outdated software.

So selective in what your taking in.

Looks like we should wait.

humbug · 21 Sep 2017 at 18:23

Doobedoo said:
So from that video it scores 100 points more running at 3.7ghz than a 1600 running at 3.2ghz.

So it would seem.

gavinh87 · 21 Sep 2017 at 18:24

Doobedoo said:
Well I'm sceptical as I'm sure you can understand. He provided a screenshot of 6 cores running at 5.2ghz. That is it!

True, we will soon see the real benches. Not long now!

Also, that video is running software from an external drive, who knows how old the software is on there?

gavinh87 · 21 Sep 2017 at 18:25

Doobedoo said:
So selective in what your taking in. Looks like we should wait.

Im just dubios of this video, its the only source of anything that shows that score being so low.
As said, PCPER have dismissed this and so has BVTA.

Doobedoo · 21 Sep 2017 at 18:25

gavinh87 said:
True, we will soon see the real benches. Not long now!

Also, that video is running software from an external drive, who knows how old the software is on there?

What sort of a performance impact does the different versions have?

Doobedoo · 21 Sep 2017 at 18:27

gavinh87 said:
Im just dubios of this video, its the only source of anything that shows that score being so low.
As said, PCPER have dismissed this and so has BVTA.

But your willing to believe a random dude with a screen shot? No information as such just 6 anonymous cores at 5.2ghz.

Maybe it's that low because as Humbug has said it's not been overclocked. Would make sense though.

gavinh87 · 21 Sep 2017 at 18:31

Doobedoo said:
What sort of a performance impact does the different versions have?

Older builds score differently. Look at his cinebench, that drive has been passed around a lot. There are results on there that you don't get on vanilla you download from the site. I'm betting its a very old build.

Doobedoo said:
But your willing to believe a random dude with a screen shot? No information as such just 6 anonymous cores at 5.2ghz.

Maybe it's that low because as Humbug has said it's not been overclocked. Would make sense though.

I believe cpumonkey, BVTA and pcper. 3 sources to 1.

Doobedoo · 21 Sep 2017 at 18:34

gavinh87 said:
Older builds score differently. Look at his cinebench, that drive has been passed around a lot. There are results on there that you don't get on vanilla you download from the site. I'm betting its a very old build.

I believe cpumonkey, BVTA and pcper. 3 sources to 1.

Ok that's up to you. when he goes into task manager windows reports the cpu at 4.37ghz. That's the all core boost is it not?

gavinh87 · 21 Sep 2017 at 18:36

Doobedoo said:
Ok that's up to you. when he goes into task manager windows reports the cpu at 4.37ghz. That's the all core boost is it not?

Yeah supposedly the all core boost is 4.3ghz

humbug · 21 Sep 2017 at 18:36

gavinh87 said:
Older builds score differently. Look at his cinebench, that drive has been passed around a lot. There are results on there that you don't get on vanilla you download from the site. I'm betting its a very old build.

I believe cpumonkey, BVTA and pcper. 3 sources to 1.

You obviously don't know that Ryan Shrout had to go back into that tweet and say he had not realised it was running at fixed clocks.

I mean did you even watch that video? its impossible to fake that, it is what it is.

PS: when Ryan Shrout says something, assume its an emotional hyperbole reaction to something or that he has the wrong end of the stick, thats what it usually ends up being.

Its an OEM machine, CPU boost is disable... cooling and all that. 3.7Ghz 1230 points.

Doobedoo · 21 Sep 2017 at 18:39

gavinh87 said:
Yeah supposedly the all core boost is 4.3ghz

So it would make sense on the face of it. It apparently scores 1230 in that video at stock with it's all core boost, and in the other leaks it scores 1500 possibly at 4.8-5.0?

Who knows not long to find out.

AndreiD · 21 Sep 2017 at 18:40

Could be an engineering sample or pre-production BIOS versions and such, there are many reasons as to why performance there might not be the same as the final product, best take it with a bag of salt.
It could also be on point, but you know, better keep some skepticism for everything based on rumors and leaks

gavinh87 · 21 Sep 2017 at 18:43

humbug said:
You obviously don't know that Ryan Shrout had to go back into that tweet and say he had not realised it was running at fixed clocks.

I mean did you even watch that video? its impossible to fake that, it is what it is.

PS: when Ryan Shrout says something, assume its an emotional hyperbole reaction to something or that he has the wrong end of the stick, thats what it usually ends up being.

Its an OEM machine, CPU boost is disable... cooling and all that. 3.7Ghz 1230 points.

I did not see that tweet but if thats locked to 3.7 then the scores make sense. So 1500 at 4.7ghz is not so unbelievable.

humbug · 21 Sep 2017 at 18:46

AndreiD said:
Could be an engineering sample or pre-production BIOS versions and such, there are many reasons as to why performance there might not be the same as the final product, best take it with a bag of salt.
It could also be on point, but you know, better keep some skepticism for everything based on rumors and leaks

Its one of these with 2 extra cores... and the numbers add up to exactly that.

humbug · 21 Sep 2017 at 18:48

gavinh87 said:
I did not see that tweet but if thats locked to 3.7 then the scores make sense. So 1500 at 4.7ghz is not so unbelievable.

1230 + 25% = 1537

3.7Ghz + 25% = 4.625Ghz take away loss in scaling, yeah... that's a score derived from 4.7Ghz.

Doobedoo · 21 Sep 2017 at 18:51

AndreiD said:
Could be an engineering sample or pre-production BIOS versions and such, there are many reasons as to why performance there might not be the same as the final product, best take it with a bag of salt.
It could also be on point, but you know, better keep some skepticism for everything based on rumors and leaks

Completely agree.

Doobedoo · 21 Sep 2017 at 18:53

humbug said:
1230 + 25% = 1537

3.7Ghz + 25% = 4.625Ghz take away loss in scaling, yeah... that's a score derived from 4.7Ghz.

Is that taking into account it's all core boost? You would have to assume it was operating at 4.2 unless turbo was disabled which it didn't appear to be.

humbug · 21 Sep 2017 at 18:57

Doobedoo said:
Is that taking into account it's all core boost? You would have to assume it was operating at 4.2 unless turbo was disabled which it didn't appear to be.

No its locked to 3.7Ghz, at leat we should hope so because if its not with that score Intel have problems, it is just 3.7Ghz, which means that 1536 score was got with 4.7Ghz, which is not bad and what we should have expected if we didn't, its consistent with the 7700K IPC.

OEM machine, Turbo off, cooling...