X79 and/or rampage IV OC'ing thread for those of us still left....

[dalekdukesboy]

It's here! Beautiful board, I hope it's as good as it looks, I have comparison shots of it and regular RIVE IV which I'll post as well there is a significant difference even with where chokes/resistors/etc are located so heatsinks are different but chipset/cpu/memory and some connections are mainly in same place but board itself is significantly different in details and hopefully "fixes" issues of regular rampage.

 

[mouacyk]

You're really gonna stretch the 1680's legs with that. Let us know how it goes.

I envy that board ultimate for x79. But I'm looking forward to an Asrock extreme 9. Haven't used asrock since q6600 on P35.

 

[dalekdukesboy]

You're really gonna stretch the 1680's legs with that. Let us know how it goes.

I envy that board ultimate for x79. But I'm looking forward to an Asrock extreme 9. Haven't used asrock since q6600 on P35.

Thanks, the RIVE IV was the board for X79 till they released this board and after having 2 RIVE IV's which both now are parts-only boards and having this platform since 2012 I hope this is the ultimate board for this platform and the last X79 board I buy. I never had Asrock considered it but ended up with EVGA couple times and now Asus with this.

Update, still using laptop but with all the windows updates and various other things done it's OK online so I didn't get as far with RIVE black as I would have if this thing still sucked so bad, the only good thing about it sucking is it would have sped me up putting X79 back together. So I got board in and had to switch backplates and put the AIO watercooler plate on and I used alcohol and cleaned cpu and AIO plate off and CPU is also installed. Case is 95% dust free it was disgusting because it was essentially an open bench before I was swapping stuff in and out so often I left side cover off plus it motherboard and power supply etc hadn't been moved/touched since 2016 so a lot of dust sitting under those things and under all those thick power cables etc. So it should be cooler running than it has been of late just due to that, and also blew out the AIO radiator it obviously was pretty nasty in few spots under the high cfm fans.
So I'm not pushing myself too fast because when I do/have I usually have done something done because I get pretty tired after hours of kneeling over a build and in a weird way I'm a perfectionist even though I will totally half ass some things aesthetically I'm a stickler for performance and getting every screw holding components in tight as possible. I'll post a few pictures tomorrow of what I've got atm and should finish the rest pretty quickly now that I've got all my prep work done, half the time was cleaning and just finding screws and pieces I needed it seems.

 

[dalekdukesboy]

I have an Asus Rampage IV with i7-4960X processor. On the Ram slot's there are 4 piece of GSkill Trident 2666 Ram's. When I trigger, the mainboard automatically adjusts processor speed to 4000 MHz. Also increases BCLK to 125 MHz.

2666 / 2400 = 1. 1108333 if I multiply this result with original processor frequency 1.110833333 x 3600 = 4000 MHz so the reason. If I put on 2800 Rams, with same equation.
2800 / 2400 = 1.6666666 x 3600 MHz = 4200 MHz, the system boots I assume.


@PolRoger you have a 4930K processor, with same equation.

2800 / 2400 = 1.666666 x 3400 MHz = 3967 MHz when mainboard opens with Klevv modules. Can you verify this?

A quick LinX test on my setup... I noticed that "daledukesboy" and "agent_x007" are getting ~200+Gflops while I'm at ~175+Gflops is that due to the two extra cores on the Xeon cpu that you all are running?

View attachment 109654

I just saw prices for 1680 v2 on ebay, those guys are mentaly ill or living in past.

305 usd for mine months ago. Seller listed at 375 but accepted the lower offer. My mobo croaked this month too. Eps connector melted on seasonic cable.

Waiting 1 month for a much better board.

Shame I was looking for someone with a similar system as my previous to run a few benches for me for the sake of curiosity

You're really gonna stretch the 1680's legs with that. Let us know how it goes.

I envy that board ultimate for x79. But I'm looking forward to an Asrock extreme 9. Haven't used asrock since q6600 on P35.

" I live.....again."

 

[FreedomEclipse]

View attachment 109835

" I live.....again."

Do you have a 1080ti to bench?

 

[dalekdukesboy]

Do you have a 1080ti to bench?

Yeah I've had that for several months I posted about it here when I got it.

 

[FreedomEclipse]

Yeah I've had that for several months I posted about it here when I got it.

Have you run heaven and ffxv benches on it? I'll im curious to know if the 1080ti bottlenecks

 

[dalekdukesboy]

Have you run heaven and ffxv benches on it? I'll im curious to know if the 1080ti bottlenecks

No I haven't but now that I've got the new motherboard and such I will be benchmarking everything including card to see what it does with the newer RIVE board. However, I pushed it to the absolute limit and almost 4.6 ghz is stable with less voltage than before.

Who knows if those two hottest core sensors are accurate or the 77-84 sensors are most accurate but either way it's not ice cold and passing with flying colors which is very encouraging for this board and anything I do with it including benches.

 

[FreedomEclipse]

No I haven't but now that I've got the new motherboard and such I will be benchmarking everything including card to see what it does with the newer RIVE board. However, I pushed it to the absolute limit and almost 4.6 ghz is stable with less voltage than before. View attachment 109836

Who knows if those two hottest core sensors are accurate or the 77-84 sensors are most accurate but either way it's not ice cold and passing with flying colors which is very encouraging for this board and anything I do with it including benches.

Looking forward to it

 

[dalekdukesboy]

Not blisteringly good but quite good minus the copy drooping a bit but that was typical of this memory I never got it much over 60k mbs on other board however I tweaked it, but the latency is very good and rest of numbers are respectable and all perfectly stable this was at 4.45 ghz the run I did with 4.6 I had the memory at stock settings so bandwidth wasn't hot but the voltage and temp wasn't either hence why it passed lol. I will do some benchmarks of video card I had forgotten to plug in case fans so I found that plug and obviously that helped with temps, I still have side of case off so it wasn't bad but I have fans in front bays of case and none on top/rear or side obviously but just those front fans make huge difference in keeping board cool as well as components so I'll probably add one fan to rear of case I got one spot open the watercooler is on top of case so yes there are fans there but all on cooler and if I get a few fans on side of case in right spots I might really be able to chill components and video card which also was a problem...

Fans were off at idle so yeah nice and quiet but card even at 2d settings doing nothing made card 55 degrees or so and more importantly it's right next to GPU/memory etc and it was just soaking into that area and also heating it up as well. So one more reason I don't like the Nvidia fan off idle feature so I just used MSI Afterburner and said fuck you and pumped the 3 fan cooler up to 89% which you can barely hear over all my case/cooler fans anyway this is NOT a silent build.

 

[baryluk]

4.5GHz is pretty good.

My i7-3930K was running at 4.2GHz for 6 years now (since day two ;D). And only about three months ago I bumped it to 4.3GHz. Works fine.

I also found out that during some cleanups over last year, I did put my memory modules wrong, and had only 2 channels working, instead of 4, and BIOS didn't really warm me (but it was indeed telling "Dual channel mode"). LOL. Fixed. It also allowed me to push memory frequency higher than previously (for some reasons I was stuck to 1866MT/s, well because I had two DIMMs per channel! After fixing the issue, I easily got 2400MT/s).

I might push it to 4.4GHz and it will probably work, but I might need to tweak more voltage stuff. I know in the past I had issues even with 4.3GHz, but recent success with some tweaks makes me belive I can do 4.4GHz.

ASRock X79 Extreme11.

 

[dalekdukesboy]

4.5GHz is pretty good.

My i7-3930K was running at 4.2GHz for 6 years now (since day two ;D). And only about three months ago I bumped it to 4.3GHz. Works fine.

I also found out that during some cleanups over last year, I did put my memory modules wrong, and had only 2 channels working, instead of 4, and BIOS didn't really warm me (but it was indeed telling "Dual channel mode"). LOL. Fixed. It also allowed me to push memory frequency higher than previously (for some reasons I was stuck to 1866MT/s, well because I had two DIMMs per channel! After fixing the issue, I easily got 2400MT/s).

I might push it to 4.4GHz and it will probably work, but I might need to tweak more voltage stuff. I know in the past I had issues even with 4.3GHz, but recent success with some tweaks makes me belive I can do 4.4GHz.

ASRock X79 Extreme11.

Yeah 39x series needs more voltage than 49x ivy and also memory won't go as far. My overclock for 8 cores with this xeon is almost the same top stable overclock for my 6 core sandy bridge chip. I haven't yet gotten a 4960x 6 core to play with but I'm sure it'd go higher than my xeon with 2 less cores to push even though the xeon was about the best binned skew at the time.

So far 132 bclk is highest stable slightly over what the stock XMP settings for memory is at which I posted above at 131.2 or so.

 

[baryluk]

132 bclk is quiet massive change. Nice.

I am running my computer 24/7, and I didn't touch bclk, but I guess that is only option on most Xeons.

 

[dalekdukesboy]

132 bclk is quiet massive change. Nice.

I am running my computer 24/7, and I didn't touch bclk, but I guess that is only option on most Xeons.

No the Xeon is totally unlocked, the bclk is necessary to change just to get certain memory speeds because unfortunately there is no other way to change them. This memory's XMP profile is what I first posted which was 131.2 so memory manufacturer just expects you have a cpu/mobo that can run it, and if you can't...you just have to lower settings manually to get it to work.

But it's technically only about a 7 mhz increase because you have straps so I just set it to the 125 mhz and the internal clock of the mobo is actually 100 the 125 is just a virtual # so you can hit certain speeds or it'd be even more limited.

This is best result thus far, I got 2 runs and on third run it was very close to passing I can tell by the number sequence when it's really unstable/off it gives you whacky numbers but it wanted to work I changed BCLK skew and seemed to help so might need to mess with few other settings or voltages to coach it to work....if not I'll do a longer run at exactly 132 and see if I can get 10 runs or so which would show it's nice and stable. Either way, pretty hard to complain, so far this board has been plug/play and doing all I could ask for it already matched and exceeded the regular RIVE I'll have to scroll back but I believe 132 was either dicey or I never quite got it stable particularly with 2800 mhz memory.

Best result thus far. 132 bclk stable which puts memory at 2816 mhz.

 

[dalekdukesboy]

Looking forward to it

 

[FreedomEclipse]

@dalekdukesboy

Hmmmmmmm -- I cant explain it. Maybe the HT along with the quad channel memory & bigger CPU cache boosts your score? I know my 1080Ti being a 'middleware' card is technically inferior to a lot of other 1080Ti when it comes to overclocking

 

[agent_x007]

@FreedomEclipse It's simple :
@dalekdukesboy has a 4,5GHz Octa Core CPU with HT enabled, while you... don't.

 

[FreedomEclipse]

@FreedomEclipse It's simple :
@dalekdukesboy has a 4,5GHz Octa Core CPU with HT enabled, while you... don't.

Oh I missed that. I thought he was benching with a Hexa core+HT In which case, that makes plenty of sense but its good to see that our GPU scores are within the same range which means no bottlenecking.

 

[dalekdukesboy]

Oh I missed that. I thought he was benching with a Hexa core+HT In which case, that makes plenty of sense but its good to see that our GPU scores are within the same range which means no bottlenecking.

This is true, but also you could just look at individual scores and see the story, my graphics score which is obviously just the video card is barely higher than yours and this is best OC I could achieve last night I was in mid 29K range on prior runs but physics and combined scores I have are significantly higher, for reasons 007 pointed out.

Also ironically I have the exact same video card as you, I have the gigabyte OC 1080ti, there was a black and white version not sure if they were different at all other than shroud color but just fyi I have the one with the white shroud.

 

[FreedomEclipse]

Also ironically I have the exact same video card as you, I have the gigabyte OC 1080ti, there was a black and white version not sure if they were different at all other than shroud color but just fyi I have the one with the white shroud.

Our cards are the same, the difference in colour is based on what the retailer could get at the time imo. some sold black some sold white.

I highly recommend getting a Accelero Xtreme IV for your card as well as the heatsinks for the VRM/Mosfets and memory. putting it on helps the card boost to almost 2Ghz when i ran the stock cooler it struggled to stay at even 1800mhz unless i turned the fans up above 60% but by that time the card sounded like a jet engine.

Im not sure if Arctic Cooling are doing more heatsinks anymore as they wouldnt allow me to buy more than one packet of them but hit up amazon for Enzotech or Alphacool VRM or VGA heatsinks. Use some thermal adhesive instead of thermal tape and you should be good.

 

[dalekdukesboy]

Our cards are the same, the difference in colour is based on what the retailer could get at the time imo. some sold black some sold white.

I highly recommend getting a Accelero Xtreme IV for your card as well as the heatsinks for the VRM/Mosfets and memory. putting it on helps the card boost to almost 2Ghz when i ran the stock cooler it struggled to stay at even 1800mhz unless i turned the fans up above 60% but by that time the card sounded like a jet engine.

Im not sure if Arctic Cooling are doing more heatsinks anymore as they wouldnt allow me to buy more than one packet of them but hit up amazon for Enzotech or Alphacool VRM or VGA heatsinks. Use some thermal adhesive instead of thermal tape and you should be good.

Glad you said that, my card stays cool if you turn the fans up but like my old and new mobo just keeping cpu/gpu cool isn't only problem if mosfets/vrm are cooking it will throttle anyway and I have a feeling you are right because my card boosts all over the place depending on the load on it and I do have a spare cooler or two here so I'll see if any fit etc or I'll get another.

 

[FreedomEclipse]

The stock cooler on it really isnt any good and one of the reasons why the card struggles to cool the card and hold boost clocks is because the stock heatsink itself makes contact with the VRMs and cools it through thermal pads... So you're not just battling with cooling the GPU itself but also the VRMs.

Upgrading to an aftermarket cooler was probably the best thing i ever did... Easily dropped 20-25'c on the GPU-Core while not having to ramp up the fans to silly levels.

Accelero Xtreme IV or Raijintek MORPHEUS II CORE -- either is fine. so its down to personal taste with which one you pick. I think the Raijintek comes with its own ram and mosfet heatsinks so you kind of nail two birds with one stone if you get it. Accelero Xtreme IV doesnt come with heatsinks... what a dumb idea

 

[agent_x007]

Hmmmmmmm -- I cant explain it. Maybe the HT along with the quad channel memory & bigger CPU cache boosts your score? I know my 1080Ti being a 'middleware' card is technically inferior to a lot of other 1080Ti when it comes to overclocking

Definitly a too slow CPU for 3DMark 11 :

^Valid : https://www.3dmark.com/3dm11/12953702
GPU was clocked around 2080MHz (and this is FE).
Those Octa Core Xeon's are seriously OP (or 3DMark 11 is simply broken at Performance settings).
@dalekdukesboy Why I got 2000 points more for Physics than you (I clocked mine at 4,4GHz, 1866MHz memory CL8.9.8) ?

 

[dalekdukesboy]

Definitly a too slow CPU for 3DMark 11 :
View attachment 110024
^Valid : https://www.3dmark.com/3dm11/12953702
GPU was clocked around 2080MHz (and this is FE).
Those Octa Core Xeon's are seriously OP (or 3DMark 11 is simply broken at Performance settings).
@dalekdukesboy Why I got 2000 points more for Physics than you (I clocked mine at 4,4GHz, 1866MHz memory CL8.9.8) ?

Also my combined score is higher than yours while the physics is lower....weird. No idea why that is but I'll have to get a better cooler on my GPU and I'll get a better idea of what this thing can do.

 

[dalekdukesboy]

Update: Also a lesson on CPU building, if you smell something, something is wrong. Even if you don't realize it's your computer since it could be a million other things in your house that can burn or even dust just burning this time of year from your furnace kicking on. Anyway, I found out why the plastic on my CPU power plugs 12v broke off in my last motherboard. I didn't think much of it I just cursed as I broke them free of that board and assumed I jammed them in too tight or I did something wrong but found it odd of all the plugs on the power supply those 2 the 4 pin and 8 pin were only ones that did that.
Long story short, they melted in the socket. So I was breaking free the melted plastic which remained in socket on old motherboard. And I was getting random shutdowns etc on last build near end before it wouldn't boot and now I've had increasing issues with this one but even at first every so often under stress it would just stop and restart. I just assumed it was reaction to a bad overclock though strange it has happened before on other builds just not so frequently.
Anyhow weird thing is, I had plugged one remaining non-broken/melted pin into this new board and had same result and was distinct smell this time whereas before if there was I didn't realize it or assumed it was something else. So I noticed all my voltages/power was always consistent on every board and power supply is a Rosewill 1200 watt so I just didn't think it was power supply itself doing something wrong, so I used the plug that has exposed metal prongs that are what contact your board and plastic around them is just for show and to line up socket and bingo, all works fine. So only thing I can figure is plastic used on those fittings was defective or not proper plastic for obvious heat you incur where power enters board.
Other thing that is odd is I've had this power supply for at least a year or more so it's new but not so new I'd suspect it had an issue, it's one of newest/up to date things in my board part of reason I bought it! Anyway now not only is it working....I got it to run with least voltage ever Linx stable...just when you think you've seen everything you could see in 15 years of building PC's! Just....odd. But I'll take it!

Oh and besides fact that it was stable set at 1.36 volts....it was dropping to only 1.328 under load and was stable. I was using 1.375-1.39 minimum before to keep 4.5 ghz stable with 2666 mhz memory which this run was. Obviously my temps were great as well with so little voltage.

So here is screenshot on first of 15 runs showing load voltage 1.328, I laughed when I saw that and just assumed it would fail on first run or soon thereafter, definitely didn't think it would make 15 full runs.