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

[gavinh87]

Its actually about the same IPC as KabyLake, KabyLake has a clock speed advantage, Ryzen has a core advantage and a huge cost advantage, and better power consumption if you're interested in such things, makes them cheaper to cool, i guess.

Yes, The Ryzen 5 is able to keep the 7700K at arms length, its not more than 10% faster in most things, the 7700K is much faster in a few, but the Ryzen 5 is also able to keep better minimums is a few other things, smooth gameplay.

Its pretty clear when you look at these performance differences you're getting for the 7700K its less than its clock speed advantage, its not that Ryzen has better IPC, its maybe 1 or 2% less if anything, what Ryzen does have is more compute threads, more than the 7700K's 8.

In a 6 core Intel thread that is not an Insult to Intel, not anymore, something else you need to come to grips with, its just fact and a welcome one given we will now get 6 core CPU's from Intel too.

Show me any article that states ryzens IPC is close to that of kaby. I posted a graph a few days back of test the stilt did over many different benchmarks and clock for clock ryzen is about haswell perf. Which is can attest to as my single core cinebench score is the same as my haswell.

I'm not hating on ryzen here but it's IPC certainly isn't the same or even within 5% of kaby. We have to look outside of games to see this to remove GPU limitations.

When coffee arrives and the core advantage disappears i expect we will see very different results.

 

[gavinh87]

custom watercooling loop with 240 rad, non delided, on a Gigabyte motherboard.
with auto OC at 4.7Ghz the board uses 1.37v making the chip stays around 75-79 Celcius during Cinebench R15.

I imagine your temps are much lower than this as you are using 1.2xv
A decent aio should be able to keep this under control *wishful thinking*

 

[Martini1991]

@humbug can you please do a bench of Cinebench R15 Single core at 3.9GHZ? Run your fastest cache speed/RAM speed and I'll do the same with my Ryzen. I'd be shocked if there was any massive difference in it.

 

[BTVA]

I imagine your temps are much lower than this as you are using 1.2xv
A decent aio should be able to keep this under control *wishful thinking*

as in tradition, it needs deliding.

 

[Journey]

@humbug can you please do a bench of Cinebench R15 Single core at 3.9GHZ? Run your fastest cache speed/RAM speed and I'll do the same with my Ryzen. I'd be shocked if there was any massive difference in it.

I just did a quick run on Cine R15 on a system I have here that I'm changing the RAM on it's R5 1600 4.0GHz with the RAM at 3333 C14 (can't got to 3466 or it has training issues) single core on CInebench is 165, the full twelve threads was 1355.

 

[jigger]

Yep!
I can't wait till the 8700k is out to put this to bed. 8700k will murder ryzen in games.

Define murder. One system running 150 FPS the other running 170?

Even if the 8700K hits 120FPS it's pretty much all irrelevant.

 

[Martini1991]

I just did a quick run on Cine R15 on a system I have here that I'm changing the RAM on it's R5 1600 4.0GHz with the RAM at 3333 C14 (can't got to 3466 or it has training issues) single core on CInebench is 165, the full twelve threads was 1355.

I got 161 at 3.9GHZ with 2933MHZ RAM on my 1700.

Need an i5 to compare though.

 

[gavinh87]

Define murder. One system running 150 FPS the other running 170?

Even if the 8700K hits 120FPS it's pretty much all irrelevant.

Its all relevant at high refresh rate gaming.

 

[jigger]

Its all relevant at high refresh rate gaming.

Not after 90 Maybe 95 for the top half a percent of the population.

 

[humbug]

I just did a quick run on Cine R15 on a system I have here that I'm changing the RAM on it's R5 1600 4.0GHz with the RAM at 3333 C14 (can't got to 3466 or it has training issues) single core on CInebench is 165, the full twelve threads was 1355.

Well, hang on i need to go into the bios and all that... brb

 

[gavinh87]

Not after 90 Maybe 95 for the top half a percent of the population.

Who cares what % it is. Most gamers i know that will be interested in this chip will already have a high refresh rate monitor.
Those on 60hz are missing out imo.

 

[gavinh87]

For science
156 single
1689 multi
3.8/3466 ram

 

[jigger]

Who cares what % it is. Most gamers i know that will be interested in this chip will already have a high refresh rate monitor.
Those on 60hz are missing out imo.

People with sense LOL

Most game engines will have fallen to bits long before you reach the capability of the system never mind the person...

 

[gavinh87]

as in tradition, it needs deliding.

Typical
Hows the IMC?

 

[AndreiD]

I'm getting 156 single and 1674 multi with my R7 1700 at 3.8Ghz with RAM at 3200 16-16-16-36.

 

[humbug]

153 @ 3.9Ghz - 2400Mhz Ram.

 

[gavinh87]

People with sense LOL

Most game engines will have fallen to bits long before you reach the capability of the system never mind the person...

Oh do shut up with this nonsense.

153 @ 3.9Ghz - 2400Mhz Ram.

So ryzen has haswell level IPC then, as I've been saying all along.

 

[AndreiD]

Overclock it and you'll have faster single than any of us, 4690K should get to 4.5Ghz.

I found my old 2600K @ 4.5Ghz Cine R15 results:

So single was slightly faster at 4.5Ghz than it is on my 3.8Ghz 1700. 4.7Ghz 2600K was about 166 single from what I recall.

 

[DarkHorizon472]

Just tried fine r15 on mine. 1700 @ 3.7 & 3466c15 ram.

Multi core : 1665
Single core : 153

 

[humbug]

Oh do shut up with this nonsense.


So ryzen has haswell level IPC then, as I've been saying all along.

You have a higher score despite being clocked lower.

Martini got 161 at the same clock speed, that's 5% higher IPC to his CPU.