The 4790k is Devils canyon I think.

 

Yes my 4770k would hit a thermal wall at 4.2ghz (not delidded).

 

Since I can get about 4.7-4.8 on the 6700k before it gets too hot as well, there is a 500-600mhz difference.

 

The 5 way optimization actually downclocked my rams.

 

Wow 4.92 with 1.341v and that's your worst 6700k? That seems pretty good. How are your temps though?

I get over 90c max temp if I run the AI Suite benchmark with AVX at that voltage.

4790k is a Haswell , realeased as Haswell refresh

4770k normal oc is 4.5-4.6ghz 4790k 4.6-4.7ghz and last a normal 6700k 4.7-4.8ghz.

But if your unlucky to find a very bad CPU the difference can be 500mhz.

The diff between my 4770k and 6700k is not even 100mhz,

But if you compare your 4770k with mine Its a diff of 600-700mhz

If we lock at OC potential between 4770k and 4790k is that is hard to find a very good 4790k, much more easy to find a really good 4770k

I agree it's quite useful. But on my system it would not go past 4.6 no matter what.

 

And I personally feel that as a experienced overclocker the bios seems much easier for me to access and change settings if I need to. The Ai Suite is good, but intended at a more mainstream audience getting into overclocking IMO. After trying it out I still feel the bios is a better and easier option for me.

 

It's a good starting point yes, but not when you still have to spend time fine tuning it afterwards whereas in the bios I can do everything I want to at one go. Asus has also made the bios much more user friendly and accessible. All the overclock specific settings are in one page. The AI Suite however, seems quite messy to me if you are looking for specific settings to tweak.

 

 

 

 

Without knowing the settings you used within 5 way, it's impossible to comment further as to why you only achieved 4.6. There are a number of settings that would limit your overclock.

 

You misunderstand when I say it's a good starting point. I meant run 5 way, go away and have a cup of tea, watch game of thrones, [awesome] allow it to run it's stress tests, [and I recommend much higher than the default 15 seconds per test, the stress tests within 5 way can be set to as high as 6 hours total]. Upon your return, if you aren't happy, you can enter the UEFI and make any minor changes. I feel though, that if set correctly that wouldn't be required.

 

 

I don't want to go to over the top, as I myself haven't built my rig yet or dived into 5 way myself.I may find I too prefer the UEFI. However, from the comments I've seen on Avsim, I still say many are underestimating 5 way, and are not aware of the function of many of the settings.

 

As for 5 way being intended for just the beginner, first getting into overclocking... Asus say not. They say the software is aimed and beginners and advanced overclockers alike.

4790k is a Haswell , realeased as Haswell refresh

4770k normal oc is 4.5-4.6ghz 4790k 4.6-4.7ghz and last a normal 6700k 4.7-4.8ghz.

But if your unlucky to find a very bad CPU the difference can be 500mhz.

The diff between my 4770k and 6700k is not even 100mhz,

But if you compare your 4770k with mine Its a diff of 600-700mhz

If we lock at OC potential between 4770k and 4790k is that is hard to find a very good 4790k, much more easy to find a really good 4770k

 

I must have been quite unlucky with my 4770k then. Unless you have yours delidded I definitely cannot go that high.

 

 

Without knowing the settings you used within 5 way, it's impossible to comment further as to why you only achieved 4.6. There are a number of settings that would limit your overclock.

 

You misunderstand when I say it's a good starting point. I meant run 5 way, go away and have a cup of tea, watch game of thrones, [awesome] allow it to run it's stress tests, [and I recommend much higher than the default 15 seconds per test, the stress tests within 5 way can be set to as high as 6 hours total]. Upon your return, if you aren't happy, you can enter the UEFI and make any minor changes. I feel though, that if set correctly that wouldn't be required.

 

 

I don't want to go to over the top, as I myself haven't built my rig yet or dived into 5 way myself.I may find I too prefer the UEFI. However, from the comments I've seen on Avsim, I still say many are underestimating 5 way, and are not aware of the function of many of the settings.

 

As for 5 way being intended for just the beginner, first getting into overclocking... Asus say not. They say the software is aimed and beginners and advanced overclockers alike.

 

I'm not sure why too, but it might be the temps that trigger the 5 way to stop at 4.6ghz. Do you tick the AVX testing as well? Without AVX the temp difference is significant. In the videos on Youtube by Asus, they do not include AVX in the stress testing which is why the temps are so low. I find that the temps I get when running P3D are much closer to the tests with AXV included. 

I can think of a couple of possibilities.

When you achieved 4.8 manually and stress tested, did you include AVX in that stress test? If not, probably why you achieved 4.8. So not surprising that 5 way only gave you 4.6 with AVX selected.

The other possibility, if you selected AVX both for your manual overclocking stress testing and for 5 way optimisation, is that 5 way runs stress tests that are different to the tests you ran. IT's possible for example to pass Prime95 with flying colours but fail Handbrake within seconds. Asus include a wide range of real world stress tests within 5 way. In other words, it may be that your 4.8 isn't truly stable and may be an issue in the future under certain conditions. Whereas the 5 way optimization's 4.6 is truly stable.

 

 

Do you tick the AVX testing as well?

 

 

 

To be honest, if you don't run synthetic stress tests with AVX, or professional applications that use AVX, it's not necessary.

 

Some enthusiasts like to stress test with AVX because they feel it represents total stability under the most stressful scenario. 

I can think of a couple of possibilities.

 

When you achieved 4.8 manually and stress tested, did you include AVX in that stress test? If not, probably why you achieved 4.8. So not surprising that 5 way only gave you 4.6 with AVX selected.

 

The other possibility, if you selected AVX both for your manual overclocking stress testing and for 5 way optimisation, is that 5 way runs stress tests that are different to the tests you ran. IT's possible for example to pass Prime95 with flying colours but fail Handbrake within seconds. Asus include a wide range of real world stress tests within 5 way. In other words, it may be that your 4.8 isn't truly stable and may be an issue in the future under certain conditions. Whereas the 5 way optimization's 4.6 is truly stable.

 

 

 

 

To be honest, if you don't run synthetic stress tests with AVX, or professional applications that use AVX, it's not necessary.

 

Some enthusiasts like to stress test with AVX because they feel it represents total stability under the most stressful scenario. 

 

The manual overclocks I did were tested with both Cinebench, Aida64 and also real world usage.

 

I've been playing around with it and I'm pretty sure the limiting factor in the 5 way is temps. As it more or less is in manual overclocking as well.

 

Like I said, the temps I get in P3D are more representative of those I get when stress testing with AVX included or vice versa. That's why I choose to include AVX. Prime95 temps are way worse, which is why stopped using it to stress test. Trying out Realbench now.

 

I seem to be having the same issue when it comes to temps with the 6700k as with the 4770k. I've been reading abit and its most likely not the watercooling loop as idle temps are around ambient +5c which is said to be good. It's only under load when temps skyrocket.

 

For example now at 4.7ghz at 1.28v I'm already hitting the 90s just with cinebench and realbench. Max temp in P3D was about 93c. There is something definitely off here.

Hi all,

 

need to reactivate this thread because i am thinking of upgrading mi i3570k that runs at 4.4 to a 6700k

 

Will upgrade Ram too and running a 970gtx with 4gbyte.

 

Do you think this processor upgrade will have more then just a few more fps for my p3d 3.2 setup ?

 

Thanks for your help, Carsten