UnOfficial EVGA X58 Classified Owners Thread

[cstkl1]

thanks a lot SAE. I've never seen my rig this stable with that cpu clock together with that uncore and ram freq. played with these voltages the whole night and was still unstable. Got pissed and turned on turbo. turned up some vcore and tweaked DIMM DQ vRef to -50 and used 1.61 vdimm which was the one that gave me the most runs b4 it crashes. I was just trying my luck and don't even really know when to adjust DIMM DQ vRef lol ...and got this.. this is just wierd =.=". Gonna let it run for one whole night and see what happens. if this is stable overnight gonna try for 7-8-7-24. Is that possible with this sticks with just air cooling on the ram?

yup that dqvref at -50..
it was actually gaygay mobos that ppl started noticing vref around 0.75v gives better stability

judging from ure vdimm if 8-8-8-24 is 1.61v set than ure 7-8-7-20 should be around 1.71

[SAE]

For everything I have measured via DMM, yes just vdimm and vtt.

Vdimm is just being rounded up to the next 100th. 1.667V via DMM is 1.67V on Eleet.
Vtt is showing about 0.1V too much for many people. Measured via DMM there's even less voltage than what was set. 1.35V BIOS (+150) equals 1.324V loaded for me, while Eleet is reading 1.44V!

[kill_a_wat]

Hehe looks like i'll need to buy a DMM to check my specific system. I am 99% sure though that the VTT is overstating volts

[blizz]

7-8-7-xx should be possible, of course. Depends on your sticks though. Aircooling is enough, my 1057 7-8-7-21 1N were on air as well (stock Corsair dual fan). You might have to up vdimm a bit to run 7-8-7-21. Try it out if stable. Big thumbs up for ya.

wierd again. 1 worker stopped after 2 hours, then another an hour later and when I re-ran the test again, it crashed within 10 minutes for all 5 tries i did.. it was a fluke =.=". RTL was dead on on 58/61/63 all 5 times too =.=. wth?

[SAE]

Hmm. You set it up using XMP with custom voltages and then you just modified the main timings?
Do you still use VREF = -50? Maybe change that, cause i never used vref successfully (should try it later though). I also could not find anything positive from using diffent skew values.

CL7 definitely stresses the IMC more. So Vtt, Vdimm, Vcore may need some lil extra voltage.

One app I found quite useful for memory testing was HCI Memtest for Windows. I was using older versions... with 8 thread you should open 8 instances of it and divide the free memory (to be tested) into 8 equal parts. So every thread (4 virtual, 4 cores) is being stressed and results are more reliable than memtest for dos as that's only using one core.
http://www.hcidesign.com/memtest/

[blizz]

Hmm. You set it up using XMP with custom voltages and then you just modified the main timings?
Do you still use VREF = -50? Maybe change that, cause i never used vref successfully (should try it later though). I also could not find anything positive from using diffent skew values.

CL7 definitely stresses the IMC more. So Vtt, Vdimm, Vcore may need some lil extra voltage.

One app I found quite useful for memory testing was HCI Memtest for Windows. I was using older versions... with 8 thread you should open 8 instances of it and divide the free memory (to be tested) into 8 equal parts. So every thread (4 virtual, 4 cores) is being stressed and results are more reliable than memtest for dos as that's only using one core.
http://www.hcidesign.com/memtest/

I was still at c8 with xmp with modified timings. Wasn't pleased with 2 hours prime with 2 workers stopped. Frustrated as hell. Thought I was closer to getting it stable but rerunning prime got me restarts within 10 mins...i'm back to square 1...
ok I'll try HCIDesign memtest. HT off = 4 instances with 5400/4?

[SAE]

I was still at c8 with xmp with modified timings. Wasn't pleased with 2 hours prime with 2 workers stopped. Frustrated as hell. Thought I was closer to getting it stable but rerunning prime got me restarts within 10 mins...i'm back to square 1...
ok I'll try HCIDesign memtest. HT off = 4 instances with 5400/4?

Weird acting system. Did you already try to change your sticks' order like:
stick 1-2-3
stick 1-3-2
stick 2-1-3
stick 2-3-1
stick 3-1-2
stick 3-2-1

Sometimes sticks prefer one slot. I already noticed that in my good old athlon xp times.

[cstkl1]

sae bro
i found the solution for 210mhz 7-8-7-20 1n 2100mhz with maintaining that two setting at 4 instead of 6

now the hard part is getting to 2150

[SAE]

Nicey. What was the problem?

[Aedubber]

DO you guys have the VDropp enabled or disabled? I OCed my setup and Vdroop is enabled with a few sttings on auto etc.

[ReverendMaynard]

DO you guys have the VDropp enabled or disabled? I OCed my setup and Vdroop is enabled with a few sttings on auto etc.

vdroop disabled here. Always used some sort of load line calibration since back on my P5K deluxe. Just seems more stable and predictable for me but some have had issues with it.

[gmat]

vdroop disabled here. Always used some sort of load line calibration since back on my P5K deluxe.

Exactly same situation (since P5K black pearl) - makes sense for me to have lower Vcore at idle.

[blizz]

Weird acting system. Did you already try to change your sticks' order like:
stick 1-2-3
stick 1-3-2
stick 2-1-3
stick 2-3-1
stick 3-1-2
stick 3-2-1

Sometimes sticks prefer one slot. I already noticed that in my good old athlon xp times.

hey sae do we have to adjust rtls or are we suppose to leave them on auto. Whats the rtl for 8-8-8-24?

[cstkl1]

Nicey. What was the problem?

will message ya. later after i do the screenshot after i confirmed its ok in 59,49 and J bios

also instead of 7-8-7-20.. solved for 7-7-7-21.. difference between both on read not much but latencies do.

so 2100mhz bclk 210 7-7-7-21 1n.

[SAE]

hey sae do we have to adjust rtls or are we suppose to leave them on auto. Whats the rtl for 8-8-8-24?

It depends on the board's auto detection routine, that sets the RTL timings according to clocks, timings and some imc feedback iirc. Funny thing is, RTL can get determined differently every boot. So there's no fixed values. For me RTL was working fine at 53-55-57 or 53-56-58 at 2000 7-7-7-20 and the higer i got, the slower i had to set them up. 1057 7-8-7-21 was like 56-58-60 then. But you may have to set them to 55-57-59 first to try if stable. Later you can tweak them more.

will message ya. later after i do the screenshot after i confirmed its ok in 59,49 and J bios

also instead of 7-8-7-20.. solved for 7-7-7-21.. difference between both on read not much but latencies do.

so 2100mhz bclk 210 7-7-7-21 1n.

Hey buddy,

thanks. Looking forward to it. I did not check out the newer bioses as they're optimized for GT/WM, yet.
I cannot do any testing atm, as there's problems with my cpu cooling. But I hope I can commence with serious clocking by the end of may.

[cstkl1]

7-7-7-21 false lol.. can pass linx 100 runs but will fail prime 95 5.4gb
shows how bad linx is on detecting dram main timings.
linx is however pretty good for the uncore/ioh subtimings

7-8-7-20 it is

[t_ski]

I just upgraded my CPU from a 920 to a 980X. To be sure to have full CPU microcode support, I updated to the latest BIOS (59 IIRC). Now that I am starting to overclock the CPU, I found a minor issue upon rebooting the computer. The system reboots, then the system powers off momentarily, the powers back on. Watching the debug LED, it scrolls through several numbers, displays "68" for a half second, then it powers off as it changes to "69."

I found this thread here:

http://www.xtremesystems.org/forums/...d.php?t=218840

...which shows this:

68 PM Init
69 Init Cache Controller

Do any of you feel that this is an instability, or is this a function that's necessary upon overclocking?

[ghost_recon88]

So which BIOS has been the best for Bloomfield subzero?

[SAE]

Board?

[blizz]

SAE bro. First time pass 2 hour prime blend 5.4gb. TRFC at 110 though will try with 88 before I sleep...RTLS are 60,61,62. I hope it doesn't go to hell after a reboot like last time.

Also if 1.66 vdimm is needed for C8 that as per CSTKL's post it probably means around 1.76 vdimm would be needed for C7. Is it safe to run it with so high vdimm with only Corsair Ram Air cooler?

[SAE]

Hmm, weird it needs this much vtt... I guess that's why the vdimm has to be 1.66V (quite high). But you don't have to worry about aircooling the ram, many guys don't cool their ram at all. The Corsair RAM fan is intended to be used with overclocking. And those hyper can stand 1.7V easily, as long as it's not from the faulty batch.

[blizz]

Hmm, weird it needs this much vtt... I guess that's why the vdimm has to be 1.66V (quite high). But you don't have to worry about aircooling the ram, many guys don't cool their ram at all. The Corsair RAM fan is intended to be used with overclocking. And those hyper can stand 1.7V easily, as long as it's not from the faulty batch.

I've tried with same ratios 1.4 vtt/1.61vdimm and can't get it stable once I raise Ram multi to 2:10. on uncore 4000 but ram multi 2:8 its stable.
does means I have crap imc or ram or crap OC settings?

[cstkl1]

hmm try this bliz

only do the timings 8-8-8-24 1n tfaw 30
set both pwm frequency to the lowest
cpu and ioh pll 1.8 set to 1.5v
ich 1.05 to 1.1v
if u have to use vtt pwm at 3xx and cpu at 9xx use cpu ske 100

try this.

also want to question u..
ure cpu 8 pin eps power are u using two or one on ure board
and if one which one??

[SAE]

Maybe the ram is the problem and not the imc... if the corsair won't run its specs, it may be time for rma. maybe you should contact the corsair reps here to ask for assistance, first. you already tested uncore capabilies independantly. They will probably ask you to test with near stock clocks and 14x ram multiplier (143mhz bclk) to be sure.

/edit. cstkl1, mate. I think it should run at xmp settings at least. those are quite conservative. he adjusted voltages and found stable ones... I think corsair's binning is superior, but maybe those rams are degrading (old hypers?).

Maybe he should try your suggestet settings and then - if no success - he should test at 21x143 14x ram xmp with reduced vtt. if it does not work out, i will need to contact the corsair reps.

[blizz]

hmm try this bliz

only do the timings 8-8-8-24 1n tfaw 30
set both pwm frequency to the lowest
cpu and qpi pll 1.8 set to 1.5v
ich 1.05 to 1.1v
if u have to use vtt pwm at 3xx and cpu at 9xx use cpu ske 100

try this.

also want to question u..
ure cpu 8 pin eps power are u using two or one on ure board
and if one which one??

awesome nice hint bro. I'll try it once I finish 6 hours prime here. At least I wanna keep one stable setting for 2k mhz ram =P

CPU power I use 1x8pin and 2x4pin coz my PSU dun have 2 8 pins means all slots occupied on mobo.