4790k oc to match 6700k

[jjohnthedon1]

How far would I have to oc my 4790k to match a 6700k at say 4.6ghz

[done12many2]

3 minutes ago, jjohnthedon1 said:

How far would I have to oc my 4790k to match a 6700k at say 4.6ghz

 

You're going to need between 4.8 and 4.9 to match it.  The IPC difference is over 200 MHz, but not quite 300 MHz.

[tarfeef101]

maybe 4.8, 4.9? Honestly not much of a difference...

[pwn_intended]

you want a 4790k to perform as well as a 6700k that is at 4.6?

 

You're gonna need liquid nitrogen for that. I'm too lazy to do the maths, but I'm gonna say over 5.0ghz at least.

[done12many2]

1 minute ago, tarfeef101 said:

maybe 4.8, 4.9? Honestly not much of a difference...

 

Great minds think alike.  

[Starelementpoke]

8 minutes ago, jjohnthedon1 said:

How far would I have to oc my 4790k to match a 6700k at say 4.6ghz

About a 5% IPC improvement from Haswell so 4.6GHz*.05=.23, so around 4.83GHz? Give or take a bit.

[Drak3]

About 4.65GHz to 4.7GHz, clock for clock, the 6700K is a piss poor improvement over the 4790K. Under a good selection of scenarios, the two chips are equals. It's optimizations in Skylake that give it an edge in some other scenarios.

[jjohnthedon1]

So my cpu is far from bottlenecking or being obsolete 

[done12many2]

1 minute ago, jjohnthedon1 said:

So my cpu is far from bottlenecking or being obsolete 

 

Absolutely.  If you are running at 4.7 GHz, your CPU is performing roughly the same as a 6700k at 4.5 or slightly lower.  You're solid.

[Hayabusa1989]

At 4.5-4.6 i get the same speeds if not slightly faster than the 6700k running at stock (4.2 Turbo) so from what i see its 300-500mhz. Not too shabby at all. 

[CostcoSamples]

The answer is not simply Skylake = 1.05*Haswell.

 

Digital Foundry has done some really great work showing how the different CPUs compare in different games. We see quite a significant difference between CPU generations, using the same GPU.  Check this out:

 

 

[jjohnthedon1]

15 minutes ago, CostcoSamples said:

The answer is not simply Skylake = 1.05*Haswell.

 

Digital Foundry has done some really great work showing how the different CPUs compare in different games. We see quite a significant difference between CPU generations, using the same GPU.  Check this out:

 

 

This is good but it's a stock 4790k which is at 4.1ghz when all cores are under full load 

[jjohnthedon1]

23 minutes ago, CostcoSamples said:

The answer is not simply Skylake = 1.05*Haswell.

 

Digital Foundry has done some really great work showing how the different CPUs compare in different games. We see quite a significant difference between CPU generations, using the same GPU.  Check this out:

 

 

What would have been nice is if they clocked all the CPUs down to 2 ghz to force a cpu bottleneck 

and then showed the difference between the architectures 

[CostcoSamples]

2 minutes ago, jjohnthedon1 said:

What would have been nice is if they clocked all the CPUs down to 2 ghz to force a cpu bottleneck 

and then showed the difference between the architectures 

yeah that would be nice.  There are a lot of ways you could tease out the differences between CPUs.  As it is, the Digital Foundry videos represent a huge undertaking to synchronize the benchmarks and show the graphs in real time.  Pretty damn awesome if you ask me!  It's kind of a shame that most of it is on cutscenes rather than actual game play.  You'd have to write a program to play the game exactly the same each time through, so I understand why he didn't do that.

[done12many2]

19 minutes ago, jjohnthedon1 said:

What would have been nice is if they clocked all the CPUs down to 2 ghz to force a cpu bottleneck 

and then showed the difference between the architectures 

 

I'm actually not too impressed with several of Digital Foundry's videos.  On more than one occasion he's compared CPUs at different clock speeds attempting to imply an advantage of one over another.  In the cases where one wouldn't clock as high as another you'd think that he would just run them all at a lower equal overclocks, but he doesn't. In the case of this particular video, why include the overclocked 6700k at all?  If you are going to compare chips, show them all at stock or show them all at the same overclocked speed.

 

In addition, in all of his videos you're left to assume that the same supporting enhancements (cache speed, mem speed and so on) have been made across all platforms.  This is a common problem with a lot of reviewers on YouTube.  We're just supposed to assume that all things were equal during testing.

[jjohnthedon1]

Ok 

3 hours ago, done12many2 said:

 

I'm actually not too impressed with several of Digital Foundry's videos.  On more than one occasion he's compared CPUs at different clock speeds attempting to imply an advantage of one over another.  In the cases where one wouldn't clock as high as another you'd think that he would just run them all at a lower equal overclocks, but he doesn't. In the case of this particular video, why include the overclocked 6700k at all?  If you are going to compare chips, show them all at stock or show them all at the same overclocked speed.

 

In addition, in all of his videos you're left to assume that the same supporting enhancements (cache speed, mem speed and so on) have been made across all platforms.  This is a common problem with a lot of reviewers on YouTube.  We're just supposed to assume that all things were equal during testing.

Ok so my cousin had a i5 6500 

same ram as me but drr4 

runningb at 2400mhZ.    

I clocked mine to the same as his and disabled hyperthreading so we're esentualy both quad cores at the same speed one haswell one skylake ran a few benches with my 980ti at 720p low to eliviate and gpu bottlenecking.

both CPUs ran at 100 percent load in shadows of mordor, gtav

and rise of the tomb raider 

and skylake beat haswell by 5-7 frames, however every test was in the high 100's / low 200's. 

Gpu load was in the 60's 

so I guess I bottlenecked the CPUs and there was no gpu bottle neck and the results were minimal,

there was about 20 fps in the farcry primal bench but 170 and 190 so nothing to really consider 

had a fun afternoon testing and basically answered my own question

skylake is better but not by a fat lot