• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

ASUS ROG MAXIMUS IX HERO with 6700k overclock?

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
After running realbench a few times, i kept noticing a white screen pop up as well saying luxmark is closing so i assumed it was crashing. I even set everything back to default and it still crashed. I reinstalled graphics driver and it passed 15 minutes with 32gb of ram at 2133mhz(stock speed).

I went to bios again and just enabled the xmp so that the ram would run at 3200mhz at 1.35v and tried running realbench again and luxmark crashed...? Anyone know why this would be happening?

I really hope this RAM set is defective too. They all read fine on boot and in bios, as well as well as CPUZ, realbench, etc, so I dont even know what it can be as all im doing is setting the ram to XMP so it runs at 3200mhz instead of 2133mhz and the voltage is set at 1.35 for the ram.
 
Last edited:
Reinstalled graphics driver again with XMP settings and it PASSED realbench for 15 minutes at 32gb of ram with no problem. I dont know why i have to keep reinstalling the graphics driver but i read it helped with someone who had the same problem with luxmark crashing. Im thinking maybe when i was overclocking the CPU, the same could have happened. Maybe if I try agaian and try installing the GPU driver before stress testing with realbench?
 
I have had problems with needing to reinstall the graphics driver for running RealBench. However I never had a RealBench crash with the cause being the graphics driver.:-/

I have found that when selecting memory usage in RealBench v2.43 select less than what you have for total memory or it will cause problems.
 
I have had problems with needing to reinstall the graphics driver for running RealBench. However I never had a RealBench crash with the cause being the graphics driver.:-/

I have found that when selecting memory usage in RealBench v2.43 select less than what you have for total memory or it will cause problems.

Gotcha. So use 16gb when running realbench. It passed fine on real bench with 32gb after reinstalling driver for the 15 min test. Never had a problem with just enabling the XMP and using all the ram speed of 3200mhz. I never had a problem with my gtx1980 on my 6700k z170 build on realbench. It ran for at least 2 hours with my GPU and with my 6700k @ 4.5ghz. I dont know if it is th graphics driver, i just know that luxmarx closes sometimes, but after reinstalling driver, it tested fine on default settings and on XMP for the RAM. So I dont know exactly whats going on. Also when stress testing, my mouse freezes every once in a while then after it will work fine again, but it did pass the test @ 32GB and the screen didnt black out at all either when they passed. Could it be because im using an hdmi to hdmi cable for my monitor? the gtx1080 is capable of handling up to a 1440p at max settings for games, which is all ill ever want. 1440p is perfect for my needs. if i wish to use 4k, i can, but to play on max settings at 4k, id probably need the gtx1080 TI. Maybe in the near future ;)

Should i try taking out the GPU and putting back in? Or uninstalling the built in GPU? It's only when using realbench that it blacks out when its not working, but once i reinstall graphic driver, its perfect. Even at stock settings, luxmark closed on me a few times and after reinstalling driver, it ran realbench just fine and passed.

The power cables are in correctly. i have 2 PCIE 6+2 cables and i use one full cable and the other i use only 6 and the other 2 cables i tucked in the back. I dont know if thats causing a problem? Not a big deal. Im gaming on max settings with no problems as well.
 
I don't know what is going on if you have to install the graphic drivers all the time.

According to A LOT of people on the different boards, Realbench causes false positives all the time and im not the only person who has had this problem. Now it wont pass more than 5 minutes at 16gb of ram, but yesterday it passed 15 minutes at 32gb of ram twice after reinstalling driver. I'm gonna work with prime instead. In the past year since ive had my gtx1080, ive never OC'd it and never had any problems with it at all. I dont know how swapping the gpu into another system would make it unstable especially with just XMP settings for the ram.

Update: checked event log and saw alot of these:

-Driver Management concluded the process to install driver nv_dispi.inf_amd64_e69a53b8ddde469c\nv_dispi.inf for Device Instance ID PCI\VEN_10DE&DEV_1B80&SUBSYS_85921043&REV_A1\4&35D4F288&0&0008 with the following status: 0x0.

-Driver Management has concluded the process to add Service nvlddmkm for Device Instance ID PCI\VEN_10DE&DEV_1B80&SUBSYS_85921043&REV_A1\4&35D4F288&0&0008 with the following status: 0.

-Device PCI\VEN_10DE&DEV_1B80&SUBSYS_85921043&REV_A1\4&35d4f288&0&0008 was migrated.

-Last Device Instance Id: PCI\VEN_10DE&DEV_1B80&SUBSYS_85921043&REV_A1\4&35d4f288&0&0008
Class Guid: {4d36e968-e325-11ce-bfc1-08002be10318}
Location Path:
Migration Rank: 0x0
Present: true

Just have a bunch of those. Can anyone elaborate for me in newbie terms, lol? Is that from reinstalling the graphic drivers or is that an error on my PC, specifically my GPU?
 
Last edited:
Update: Using Prime95 now asNvidia says i have the latest driver and i dont see why they need to stress test my gpu as i havent even overclocked it. Anyways, started with 1.28v @ 4.8ghz, running prime95 latest version, blend test. Half an hour passed! Temps in high 70s, hitting 80C at the most. I dont know how I got 1.28v, i just figured id start at that, and so far so good. If it lasts a full hour, I will try 4.9ghz at the same voltage and i will try to hit 5ghz today and if i get a stable OC, i will stress test it overnight and for 24 hours straight. I will not go above 1.35v with prime95 as recommended by many, 1.40 if youre using realbench.

All I did was keep the XMP settings as well as the X4 settings for my M.2 drive, changed core ratio to 48 from auto, change voltage to manual and set it to 1.28. So basically just 2 settings, the core ratio and core voltage, thats it.

Took today off from work, so I got all day to finally get my new and first build overclocked :D

And I might sell my Asus Strix GTX1080 to a buddy's friend for $350 and get the new 1080ti since i got about 260 bucks in credit on amazon, id only need to put out about 100-200 bucks out of pocket to upgrade to the TI :D. If not, ill stick with my gtx1080, like i said, it an handle 1440p at max settings so thats perfect for me.

So far, im pretty impressed with the CPU and I just started.

Regards,

-Jordan

Update: At 4.9ghz and 1.31v, it seems pretty good so far, but the temps on HWMonitor concerns me. It shows 90C at the max while Asus AIsuite only shows 78C. I figured with all the sensors on the motherboards nowadays, the AIsuite would be more accurate but i figured id ask anyway. Also CPUZ shows 1.25vcore so i know thats not accurate. I remember you saying the 7700k can go up to 90C, and i figured if it hit 90 at 1.32v already, then if i up it to try and hit 5ghz, the temps would be close to 100c or over according to HWMonitor but on Asus suite, it would be high 80s to 90C. Suggestions on which temp to trust? Obviously i saw this coming as prime runs 15-20 degrees celcius hotter, but i forgot about the different temp readings from differnt apps.

Got me thinking, maybe I should send it to silicon lottery for their delid service. 15-25 degrees is a big difference and would actually make the processor last longer IMO, since there will be less heat going through it and better IHS inside the CPU. Has anyone had experience with silicon lottery? Ive spoke to them and was going to do the same for my 6700k, and they had a point when they said since the CPU will be running cooler, it "should" make the CPU have a longer life but its more like having a cpu for 14 years compared to 15 years, or something like that. That's what they told me when i was going to send them my 6700k, but back then, i didnt want to start taking apart the custom PC that was built for me and void the warranty, but now i know how to put it back together, so im glad.

I also just ordered some Thermal Grizzly Kryonaut thermal paste for my 7700k build, as I have whatever thermal paste that the thermaltake 360 riing PWM AIO came with, so im gonna clean the old stuff off with some alcohol swabs and put a nice drop of the new thermal paste in between the cpu and cooler block, it should yield 2-3 degrees cooler, but im sure its better than what it came with.

Update: 4.9ghz at 1.31v ran for 1 hour and 15 minutes and then the last core(hyperthread/digitalcore) stopped working. Will try 1.32v, starting to worry that it wont hit 5ghz unless i use more than 1.35v which is not recomended according to the guide on ASUS's website unless I get the 7700k delidded.
 
Last edited:
Asus Strix GTX1080 will be gone by the end of today for $300. Seems like ill be getting a GTX1080 TI :D

Anyone have suggestions as to which aftermarket one is a good one? I heard the Asus Strix GTX1080 Ti OC Edition isnt that great for overclocking. I hear the EVGA gtx1080 ti FTW3 is really good but out of stock on amazon with no date as to when it will return. Im thinking Gigabyte gtx1080ti Aurus for $720 or Gigabyte gtx1080ti Gaming OC for $700. I prefer Asus and Gigabyte components when I can. I'm sure whatever 1080ti i get, i will love it, but in terms of overclocking, apparently the 1080ti can overclock to 20-25% more speed which is ALOT. Even founders edition 1080ti's overclock to 20-25% more speed from what ive read.
 
Last edited:
From what I read it looks like you are doing fine.:) If you want to find out what temperature reading is correct the CPU will throttle the clock speed at 100c to prevent overheating, so you can check for that. For video cards I like Gigabyte and EVGA. Do you have a reference link for deliding when going above 1.35v?
 
From what I read it looks like you are doing fine.:) If you want to find out what temperature reading is correct the CPU will throttle the clock speed at 100c to prevent overheating, so you can check for that. For video cards I like Gigabyte and EVGA. Do you have a reference link for deliding when going above 1.35v?

https://rog.asus.com/articles/guides/the-kaby-lake-overclocking-guide/

In that guide, it says not to go over 1.35v and also If the CPU has not been de-lidded/re-lidded for a thermal paste upgrade, you’re likely to run out of thermal headroom around that voltage anyway.

Ran 2 hours straight of prime95 at 4.9ghz at 1.320v. I tried 5ghz but even at 1.365v, it wouldnt last more than 5-10 minutes tops on prime. The 7700k is 4.2ghz but with the turbo, it goes to 4.5ghz fine so i figured half a ghz more wouldnt be a big problem.

And thanks, I do believe HWMonitor is wrong on the cpu temp, but i guess ill find out because on hwmonitor, it was getting close to 100C but on Asus aisuite, it was in the 80s. So i can go above 1.35v with prime95? I really want to get 5ghz out of this. At 1.32v, i got 4.9ghz stable for 2+ hours so far. Judging from that, do you think i can hit 5ghz with my setup?

I think im gonna go with the Gigabyte for the GTX1080Ti, thanks for your suggestion.

Update: 3hours and 10 minutes, and 4.9ghz at 1.320v is still doing good. Aisuite showing temps around 71C, HWmonitor shows it went to 94C? I'm starting to think the aisuite is more accurate because 1.32v isnt that much, is it? And hwmonitor says it hit 94 degrees celcius at one point. I think its off by 10-12 degrees. I think 1.37-1.38v would allow me to hit 5ghz, I HOPE it does at least. Im gonna keep running the prime at 4.9ghz so at least I have a decent OC for it with XMP enabled, and ill save the profile just incase 5ghz doesnt work.
 
Last edited:
Hey wingman, how do i turn of the avx thign again on prime so my cpu can run cooler? My buddy's 5ghz 7700k, he told me to use an older version of prime, he doesnt know which one though, his buddy helped him. But according to him, older versions of prime run cooler and still stress test fine without the additional 10-20 degrees.

And thanks again, youre right about 100c being the throttle temp or "tj max" as some call it. That goes the same for my 6700k.
 
Looks like your doing very good with the Processor at 4.9GHz.:) As of 2/22/17, the top 59% of tested 7700Ks were able to hit 5.0GHz or greater 1.376V CPU VCORE (Or less) . LINK: https://siliconlottery.com/collections/lga-1151/products/7700k50g

I think i can hit 5ghz at 1.37-1.38vcore, just worried about temps :/

After 6 and a half hours, 1 of the 8 threads stopped working. so i went from 1.320 to 1.325. I think that should do it. I wanna get a 24 hour stable prime results. I can honestly care less about the extra 100mhz, it's just about hitting the big 5GHZ if ya know what i mean LOL.

If i delid the 7700k, getting t0 5ghz should be easy and temps would be a lot cooler obviously. Ill think about it, like i said, i can care less about 100mhz, but its just I know i can hit 5ghz, but the temps caused by prime95's stress test is my only concern as i dont want to melt the internal paste in the CPU or worse, crash a $350 processor :(

Thanks wingman, couldnt have done it without you.

Did some more research and it seems prime v26.6 is the one that doesnt give all that extra heat. Will update and see if thats true.
 
Last edited:
Prime v26.6 is DEFINITELY better. Im getting 64c on asus suite and 75C on hwmonitor right now at 4.9ghz and 1.325v. I trust asussuite temps, not because it lower but because of all the new temp sensors they put on the newer motherboards and hwmonitor is known for having inaccurate temps, as cpuZ is known for having inaccurate voltages. It says im at 1.25v on cpuz but im really on 1.325v.

Vcore on asussuite and hwmonitor is both the same, which says 1.344, but I know its on 1.325 as i set it and its on "manual" mode.

Wish me luck :)

- - - Updated - - -

Heres the link for anyone who wants the prime95 v26.6:

http://overclocking.guide/download/prime-95-v26-6-64-bit/

Update: Hasnt even hit 70C yet on asussuite and hasnt hit 80 yet on hwmonitor :) 5GHZ, YOU'RE MINE!!!!!!!! Gonna run 4.9ghz at 1.325 overnight and check what the highest temp was on hwmonitor, and subtract 10c. That should be the accurate temp. Then ill try 5ghz tomorrow and I'll go up to 1.370-1.38v at the most as long as the temps dont go over 90 on hwmonitor and 80 on asussuite.
 
Last edited:
Vcore on asussuite and hwmonitor is both the same, which says 1.344, but I know its on 1.325 as i set it and its on "manual" mode.

Where you set it and where it is can quite easily be different. Mine is set at 1.4v and shows 1.380v in HWmonitor (min 1.344v) and CPU-Z. I could be off base, it seems that where you set it is what the power section (is supposed to) sends out, the sensors read what gets there, The difference is sorted out with the LLC, or Vdroop.
 
Where you set it and where it is can quite easily be different. Mine is set at 1.4v and shows 1.380v in HWmonitor (min 1.344v) and CPU-Z. I could be off base, it seems that where you set it is what the power section (is supposed to) sends out, the sensors read what gets there, The difference is sorted out with the LLC, or Vdroop.

Yeah, I knew it was something like that. Was just saying how CPUZ's Vcore wasnt even close to accurate lol. Both hwmonitor and asussuite's voltage is exactly the same. Thanks bud.

I hope you have good luck hitting 5.0GHz.:drool:

Thanks wingman, 1.5 hours in and still stable on prime95 v26.6. Temps still havent gone above 80c on hwmonitor and 70c on asussuite :D Very happy. Wanna sleep but the noctua fans are pretty loud when theyre at 2500-3000rpms LOL.

It;s louder than my portable AC unit in my room!!! The guy that was supposed to call me for my gtx1080 never called so im gonna put it on ebay. A used one is going for $405 right now at the highest bid for my exact GPU with box, but i dont have box, so I think ill get more than about 350-400 for mine. I figured id hook my buddy's friend up and save me the hassle of posting on ebay, but whatever. It works 100% mechanically and is 100% cosmetically flawless.

I want the EVGA FTW3 1080 Ti, its $750 and a really good card from what i have read from reviews. Either that or im gonna get the Gigabyte gtx1080ti, just havent decided which. The gigabyte aftermarket 1080TIs will be back in stock on the 25th.

This is off topic but it's what I'm thinking about for my 6700k PC. LOL. For the time being, ill use my 6700k build and maybe start making that a better PC, such as a better mobo like an Asus Hero VIII since its a 6700k. I dont know if the 6700k will run on the hero IX with the latest bios, and I would prefer to have a newer bios sometimes when they have fixes or better stability, etc. anyone running a z270 with a 6700k? Can you update to the newest bios?.

2 hours stable now. Time for bed. Later guys, enjoy your weekend.
 
Last edited:
7 hours of testing, still good to go. Temps on HWMonitor is 76C, AsusSuite is 65C. Max temps HWMonitor hit was 82C, so on AsusSuite, temps were between 70-72C at the highest :) Now that sounds more like it :) On the Asus website on getting 5ghz with a dual radiator AIO, temps reached 72C and passed a 2 hour Realbench test, so these temps are more realistic and definitely takes that extra heat off that newer versions have for avx apps which no games or really any apps that I know of uses. Yes, obviously with every newer version of prime, the overclock will be MORE stable, but like i said, ive been using prime95 since my first custom built PC that my friend helped me with and after overclocking it which took me about 2 weeks straight since i was just learning how to overclock, and once stressed tested with prime, i bever had a problem. My 6700k was custom built from a guy and he uses the newer stress tests nowadays that dont take as much time as prime. He overclocked it to 4.6ghz, stable, according to him. As soon as i ran prime95, it crashed. So I honestly dont believe that stable. Im gonna stick with what i know and what has worked for me since i ever started overclocking. I bet i can probaly get the 6700k to 4.6-47ghz stable with prime95. According to the kaby lake overclocking guide, prime95 is one of the toughest brutal stress test, realbench being "medium".

Also downloaded HWInfo, the latest version and that is a lot more accurate as it matches up pretty well with what the reading are on Asussuite. I had a pretty old version of HWmonitor. So seems like the asussuite temp sensors work.

I'll keep prime running at 4.9ghz @ 1.325v until I get 24 hours straight. I consider that pretty stable. 72 hours straight is what ive done in the past for my q6600 and q9550.
 
Last edited:
Sounds like good testing. Prime95 is what I use and it does take the same amount of time to test as other stress tests, if you want it to. Are you still going for 5.0GHz?
 
Last edited:
Back