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

[dalekdukesboy]

Now this is starting to get interesting.....so close to 4.7 ghz and almost to 134 fsb! Never gotten here before, so I may partially be able to thank the RIVE black for this.

Just FYI to Mouacyk and anyone here who has my processor I have LLC set to ultra high so the actual voltage shown here is lower than load unfortunately, load voltage I believe was 1.456 volts. I mention this partially because Mouacyk was dead on right saying that is roughly what these 1680's seem to need around 4.7 volts. And as you can see my temps are "OK" but I got the one core that nudged 89 degrees so 90 is highest I want to go but processor has stayed stable right up to 94 so this is worst case scenario nothing I ever do will get it this hot so I'll also see if any VTT/VCCSA or other voltages can be nudged down even if it shaves a little off temp it'd be good plus less stress on the mobo and capacitors etc.

 

[dalekdukesboy]

Put the Klevv single sided memory back in just to see what it could do and got highest FSB and memory speed yet. I honestly didn't think it would go this far but was playing around with it yesterday and realized it had more left in the tank than I previously thought.

 

[dalekdukesboy]

Still tuning/playing around, some unsuccessful restarts but appears this cpu and the Rive Black really likes high fsb and high speed memory, tight timings and high multipliers....not so much.

 

[dalekdukesboy]

Holy shit. So close to 4.7 ghz stable with 138 fsb! Tweaked some secondary voltages and got them lower which allowed me to push furthest to date...I also elected to go with tight CR1 timings @ 2576 mhz on memory so I may see if I can get 2900+ like above screenshot at same speed because that if stable will get some nice memory #'s straight across the board.

 

[dalekdukesboy]

Current state of my pc, side panel been off for a while but had top on and one rad in case for majority of time, however I have a few watercooling units I've messed around with over last year or two and latest is the one on top of the pc which has a D5 pump built into the AOC and it so happened to be same brand as the big rad I have on floor in picture already so I decided to put the two together in a loop. The prior few stress tests I was doing with only the 240 mm rad with d5 and I can tell you temps have gone down considerably with the second rad I just need to figure out fan and rad configuration in box so it all works and is secure as well as has optimal airflow.

 

[15th Warlock]

My X79 was retired a couple of years ago, it rests in it's original retail box, along with my 3930K.

In it's prime, I was able to OC it to 5GHz, you can read about it here:

https://www.techpowerup.com/forums/...xtreme-3-way-gtx-580-sli.155527/#post-2465920

Can't believe I posted that in 2011! Thanks for the trip down nostalgia lane, glad to see some RIVE boards still rocking!

 

[dalekdukesboy]

My X79 was retired a couple of years ago, it rests in it's original retail box, along with my 3930K.

In it's prime, I was able to OC it to 5GHz, you can read about it here:

https://www.techpowerup.com/forums/...xtreme-3-way-gtx-580-sli.155527/#post-2465920

Can't believe I posted that in 2011! Thanks for the trip down nostalgia lane, glad to see some RIVE boards still rocking!

Yeah I hear you, I originally bought my X79 at a down time in my life as a way to buy myself some happiness so to speak lol. That was back in summer of 2012 it was still relatively new tech then, I still have original RIVE board I had which had bent pins and now I have a RIVE black I bought recently to replace the 2nd RIVE I had which is repairable but had a PSU issue so I just went with the black to get best possible board for my setup, and I had an ES 3960 for longest time then I had 3930k for a fair while after that came with 2nd board and now since last April I got the 8 core Xeon 1680 which is about best I can get for this board.

That mess I'm setting up there is to try and get temps down enough on 8 cores to get to 4.7 ghz with 2400+ memory and good bandwidth, so far temps came down but still won't test stable so tells me that I may be nearing limit of what can be down with above ambient cooling. But I'm going to put it back together slap as many fans as I can on those two rads and get max airflow and water flow with pump and see what happens.

 

[phill]

My X79 was retired a couple of years ago, it rests in it's original retail box, along with my 3930K.

In it's prime, I was able to OC it to 5GHz, you can read about it here:

https://www.techpowerup.com/forums/...xtreme-3-way-gtx-580-sli.155527/#post-2465920

Can't believe I posted that in 2011! Thanks for the trip down nostalgia lane, glad to see some RIVE boards still rocking!

Reminds me of my 920 D0 and Tri SLI GTX 580's as well Loved that setup... I should have moved on to X79 but never did.. The 3930k's were amazing CPUs, had a 4960X for a while back along, that did pretty well Always said if I was going to upgrade properly from the 920 I'd get an 8 core and then along came the 5960X...

Still have a X79 board, got my eyes on a Rampage 4 Black Edition if my mate ever gives it up... lol

 

[15th Warlock]

Yeah I hear you, I originally bought my X79 at a down time in my life as a way to buy myself some happiness so to speak lol. That was back in summer of 2012 it was still relatively new tech then, I still have original RIVE board I had which had bent pins and now I have a RIVE black I bought recently to replace the 2nd RIVE I had which is repairable but had a PSU issue so I just went with the black to get best possible board for my setup, and I had an ES 3960 for longest time then I had 3930k for a fair while after that came with 2nd board and now since last April I got the 8 core Xeon 1680 which is about best I can get for this board.

That mess I'm setting up there is to try and get temps down enough on 8 cores to get to 4.7 ghz with 2400+ memory and good bandwidth, so far temps came down but still won't test stable so tells me that I may be nearing limit of what can be down with above ambient cooling. But I'm going to put it back together slap as many fans as I can on those two rads and get max airflow and water flow with pump and see what happens.

Yes, Sandy Bridge was such a fun architecture to work with, so much potential, I mean, to think people were able to OC to 5GHz 8 years ago?

And then the performance uplift from the previous Core generation was just ridiculous, you don't see that anymore, made you a true believer in Moore's law huh? The sky was the limit...

Sounds like you're constrained by high temperature, using a push/pull configuration for your radiator should definitely help shave off a couple more degrees, and as a last resort, you can try dumping that radiator in a bucket with ice water, that should let you push to 5GHz.

Definitely much harder to pull such high clocks with 8 cores, good luck!

 

[dalekdukesboy]

I was going to mention you had 6 cores and I have 8 but you already did, also I'm not just trying to run benchies I'm running linx/burntest which is a torture test/stability test lol. I just got 4.7 ghz stable for 5 runs earlier today also put it in another thread. I've had this chip since I think last April and now putting that 2 rad loop together with a real d5 pump this is what I got.

However temps are NOT the problem now which is what was frustrating. I can keep it in low/mid 80's which is diddly and that's only if you believe the diodes on the 2 hottest cores. At this point what happens is it will just shut off and restart if I push it too far but it passes a couple of IBT runs. So the cpu/mobo literally just seems to "give up".

 

[mouacyk]

I was going to mention you had 6 cores and I have 8 but you already did, also I'm not just trying to run benchies I'm running linx/burntest which is a torture test/stability test lol. I just got 4.7 ghz stable for 5 runs earlier today also put it in another thread. I've had this chip since I think last April and now putting that 2 rad loop together with a real d5 pump this is what I got.

However temps are NOT the problem now which is what was frustrating. I can keep it in low/mid 80's which is diddly and that's only if you believe the diodes on the 2 hottest cores. At this point what happens is it will just shut off and restart if I push it too far but it passes a couple of IBT runs. So the cpu/mobo literally just seems to "give up".

View attachment 119430

Be careful of the restarts -- I think it's due to unclean power if it is sufficient. This happened to my 1660v2 (clocked to 4.7GHz) and eventually, it could not run any load -- it would just shut off after starting up. I discovered that my EPS connector on the motherboard had melted and possibly made a few pins go bad, lowering the power delivery considerably, enough to de-stabilize the computer even at stock. My PSU (Seasonic X850) turned out to be fine after I replaced the motherboard and used a 1680v, so I'm fairly sure it was the old motherboard connector going bad.

If you have multiple EPS connectors, use them all. I would point extra fans at the connector(s) too if you have spares, especially for stress testing.

 

[dalekdukesboy]

Yeah I have every power source plugged in and I do have fans on those connectors and the VRM which is in the same area thankfully. Without the fan/s I shudder to think. I've played around with a few things settings wise as well as made sure all connections are solid. I've actually raised the overclock a bit over what you see above and no restarts etc so may have been a bad connection or setting. I actually got it to pass 5 runs of linx at 4.720 ghz then it just failed the test but yeah something definitely wasn't quite right with power but seems good now and I'm keeping an eye on it.

 

[mouacyk]

It doesn't hurt to check your CMOS battery too, to make sure it's over 3v, because I've had restart issues with bad batteries that read under 3v.

 

[dalekdukesboy]

It doesn't hurt to check your CMOS battery too, to make sure it's over 3v, because I've had restart issues with bad batteries that read under 3v.

Thanks, did that not long ago and put new one in.

So now all I have left to do is see how far I can push it stably and also play around with memory, I might put microns back in and see if I can get them to cooperate with it at 4.7 and get 70mb/s across the board again.

 

[dalekdukesboy]

I still have the issue whatever it is, I had voltage bit higher than in my 5 run successful run I posted earlier but otherwise settings were identical and I went for 10 runs @ 4.711 I did have fsb slightly higher just noticed but anyway it went 7-8 runs successfully and restarted. So it's just odd that it doesn't fail early on or randomly I'd think if power delivery were an issue it would be more random and fail on 1st run or 2nd run or anywhere inbetween no? There is another power connector I believe for gpu and/or memory but I'll have to look at schematic and figure it out.

 

[mouacyk]

I still have the issue whatever it is, I had voltage bit higher than in my 5 run successful run I posted earlier but otherwise settings were identical and I went for 10 runs @ 4.711 I did have fsb slightly higher just noticed but anyway it went 7-8 runs successfully and restarted. So it's just odd that it doesn't fail early on or randomly I'd think if power delivery were an issue it would be more random and fail on 1st run or 2nd run or anywhere inbetween no? There is another power connector I believe for gpu and/or memory but I'll have to look at schematic and figure it out.

Some old threads suggest it's potentially a defective power phase in some batches:
https://rog.asus.com/forum/showthre...-Black-Edition-random-restart-shutdown-reboot
https://rog.asus.com/forum/showthread.php?46110-R4E-Black-UNBELIEVABLE!-DISAPPOINTED!-*Please-Read*

If you aren't already using it, try using 170% CPU capability which masks the problem and allows you to run even longer before the random shutdown happens. Other suggestions say to relax RAM timings, which might make sense in your case since you are pushing CPU further and further. CPU and RAM usually trade off against each other.

 

[dalekdukesboy]

Good info, I hope not. I haven't literally had anything fall off board but could still be a faulty phase. I already use full cpu capability etc, however my shutdowns aren't really "random" because they always happen when stress is on the board, random would be if it could happen at anytime idling or just surfing web etc.

 

[phill]

I hope these are ok here...

Found one of the EVGA Classifieds I was after

 

[Aquinus]

I hope these are ok here...

It looks, feels, and tastes like X79, so I'd say it's okay.

That's also a massive cooler. I can't stand covering DRAM slots. That stuff drives me insane.

 

[phill]

That's alright then I've got a Rampage Formula up in the loft with my 3820 (the one that's not dead that I found out! ) and I'm still waiting on a Rampage Black Edition....

It's a Noctua D14 something or other... I only used it as it was quick and easy to put on the CPU with a bit of paste to test the boards worked I'm not a fan of such massive coolers but it does manage to do the job well

Not quite sure what to do with that one yet.. If I can find a X99 Classified and I can afford at some point a X299 CPU, maybe I could get all my Classifieds up and running... Still, are there any other Xeon's I could try in my X79's?? That don't cost a bomb??

 

[dalekdukesboy]

I like it, the board is very clean and simple looking.

 

[dalekdukesboy]

Update to my issues with RIVE black....I'm typing from same rig but with a regular RIVE I ordered when I first had an issue with this board early on but thought board was OK so I never used it and decided to keep it and not sell it/return it. So, I'm glad I did lol. I put it all together yesterday but temps were high, waterblock wasn't tight, simple solution. Also I was having trouble getting it to boot....humorous reason is I was tired and I kept setting memory timings to 10-12-12-12 expecting it to work lol.

Anyway still monkeying with settings etc but reason I decided to do this was Tuesday night I clocked down the RIVE black in hopes of just living with whatever the issue was...worked for a bit but then screen starting randomly shutting on/off and it was something I had noticed on occasion before but rarely and forgot about it. However it was to point it was unusable every few seconds it turned on/off, I feared GPU could have issue but now in new mobo none of the weirdness of screen or the shutdowns happens. Now it does what you'd expect and if I have bad settings Windows crashes and gives error.

So no idea how or what is stable with this board still messing with it but it looked almost new when I got it and has newest bios and everything appears fine and the compiling issues of the black board are gone, also I notice I don't have to set the LLC as high to get high voltage delivered to board so that definitely points to a power delivery problem that was goofing everything up including my GPU had issues overclocking for a while, now I suspect it wasn't getting consistent power along with CPU/memory/etc. So I'm back to the original RIVE with 1680 V2 etc, let's see what old girl can do.

 

[mouacyk]

Update to my issues with RIVE black....I'm typing from same rig but with a regular RIVE I ordered when I first had an issue with this board early on but thought board was OK so I never used it and decided to keep it and not sell it/return it. So, I'm glad I did lol. I put it all together yesterday but temps were high, waterblock wasn't tight, simple solution. Also I was having trouble getting it to boot....humorous reason is I was tired and I kept setting memory timings to 10-12-12-12 expecting it to work lol.

Anyway still monkeying with settings etc but reason I decided to do this was Tuesday night I clocked down the RIVE black in hopes of just living with whatever the issue was...worked for a bit but then screen starting randomly shutting on/off and it was something I had noticed on occasion before but rarely and forgot about it. However it was to point it was unusable every few seconds it turned on/off, I feared GPU could have issue but now in new mobo none of the weirdness of screen or the shutdowns happens. Now it does what you'd expect and if I have bad settings Windows crashes and gives error.

So no idea how or what is stable with this board still messing with it but it looked almost new when I got it and has newest bios and everything appears fine and the compiling issues of the black board are gone, also I notice I don't have to set the LLC as high to get high voltage delivered to board so that definitely points to a power delivery problem that was goofing everything up including my GPU had issues overclocking for a while, now I suspect it wasn't getting consistent power along with CPU/memory/etc. So I'm back to the original RIVE with 1680 V2 etc, let's see what old girl can do.

That's a lot of effort to debug the issue, and it seems might just be a bad board. Back in the days, there were some activity on OCN about X79 socket overvolting and killing chips (due to the extra overvolt pin). ASUS may have done a bit too much on this platform with their high-end boards and didn't QC them properly, based on the RIVBE problems. Not sure if that's prior to their migration to full automation or after.

 

[dalekdukesboy]

That's a lot of effort to debug the issue, and it seems might just be a bad board. Back in the days, there were some activity on OCN about X79 socket overvolting and killing chips (due to the extra overvolt pin). ASUS may have done a bit too much on this platform with their high-end boards and didn't QC them properly, based on the RIVBE problems. Not sure if that's prior to their migration to full automation or after.

Yeah, trust me I wasn't at all looking forward to just ditching the board but things were just getting worse, particularly once the display just randomly blinks on and off almost constantly that was it, it's been kinda quirky and odd from the start so it probably always had a problem and me using it and pushing it just exposed it more, oh well. So I'm back to the board I started with lol, 3rd one to be precise, but considering I bought first one in 2012 that isn't so bad. I'd have loved to have the black work it seemed like a killer board it pushed memory and fsb etc to the hilt but obviously it had some circuitry issues.

Now I have a board that never just shuts off and if it crashes it's because windows fails and that is to be expected when overclocking. So far it works well but it is really picky about memory settings. If I keep this system any length of time maybe I'll buy another RIVE black down the road and give it a 2nd chance and see if I get a better board. For now, I just have to be happy I so happened to end up with an extra board I didn't intend on and all I had to do was go in other room bring it in here and do the surgery and it was done, no searching, ordering and waiting for a board to arrive. Looking on the bright side lol

 

[Aquinus]

That's a lot of effort to debug the issue, and it seems might just be a bad board. Back in the days, there were some activity on OCN about X79 socket overvolting and killing chips (due to the extra overvolt pin). ASUS may have done a bit too much on this platform with their high-end boards and didn't QC them properly, based on the RIVBE problems. Not sure if that's prior to their migration to full automation or after.

It could just be a design thing with that particular board. I have two P9X79 Deluxe boards, my old 3820 is in the other machine and the more recent 3930k in the one I'm using now. I've pushed the both pretty hard at times, but I never fried components on either board.