Thank you for your response. :) I had thought that the digital pwm was the superior one on paper.
Printable View
**** update
It wasn't LLC after all it was having an ioh core volt 1.2 or more
- while my machine runs fine while on at 4.2/bclk200 a high ioh causes COLD boot failures on my rig- atr least with my current settings.
just put it down to 1.16 5x cold boot success and 10x linx run stable (4.2 ghz) 200x21 (vcore 1.3 LLC lvl 2 ON)
trying at 1.12 ioh and seeing if its still stable- hopefully it wont be much of an issue- i notice that many people have stable 4.2 ghz on air in this thread with ioh at stock
let me know your thoughts- i wonder if some other people are affected by this- given the ealier reply form the chap who says multiple boot failures are normal behavious
cheers
Tim
Thats interesting about the ioh voltage, I havnt come accross that issue myself but it's good to have as much info as possible.
I regards to my previous post, I think you may have misunderstood me, I just meant that these boards take a long time to post when overclocked, and you can hear a slight 'tick' usually 3 times which implies that it attempts to boot 3 times before it actually "gets there".
I think perhaps the bios simply starts at lower clocks or voltages and raises them gradually to ensure a successful boot.
hmm it seems very sensitive
1.12 ioh passed linx (18 passes) but once again cold boot fails.
I have never had a cold boot failure at this clock speed (4.2) at ioh = 1.16,so i suspect the failures are for different reasons (too low - cold boot fail, too high = cold boot fail for different reason)
thanks foryour reply above BTW
i can hear the 3-4 ticks on successful boots.
on the failed boots (from cold) the whole rig powers off entirely!
EDIT:
COLD BOOTING TABLE FOR IOH (linx and prime stale while warm)
for bclk 200 +
1.12v fail
1.14v pass (x 5 tests)
1.16v pass (x 15 tests)
1.18 fail
1.20v fail
[auto] pass (5 tests)
for bclk 133 (stock)
1.10v pass
1.20v fail
wheni say fail it might pass one or two boots but if it fails after any number of tests I call it a fail
If it powers off, then comes back on a few moments later, and loops like this, then I thinkit's most likley to be RAM related, either the timings or the mem ratio, or perhaps related to uncore speed/voltage.
no its definately the IOH
evidence: tried with ioh core values as above with ram masively underclocked and loose timings.
passes linx and prime for many hours remember. also 16 passes of memtest normal
as i understand it IOH is needed for the memory and the cpu to talk to each other
i just tested it on [auto] at blck200 and it cold booted every time (5 trials) can anyone tell me how it behaves on auto?
Well, I have all but given up on hitting 4.6 on my 920 :(
I have been rock solid 24/7 and gaming for hours on end (bfbc2) @ 4.5 but it seems every setting I try to get 4.6 ends up in a couple failed boots and then me needing to reformat my raid array. I think my chip just can't handle the high blk.
can someone help me to stabilize my Dynamic Vcore settings?
My stock vcore is 1.26875v and I currently use 1.27500[bios] with LLC 1 Enabled which gives me 1.248v idle/load in CPU-Z, I want to use the DVID to lower my idle vcore, so I used the guide by setting Vcore to Normal and adding the value I want in DVID so it should be my [Normal vcore] 1.26875v + 0.00625 [DVID] which will give me 1.27500v, I get correctly the load vcore I want 1.248v but my idle vcore is way down to 0.944v which make my PC to freeze in idle just on windows screen, I want to get like 1.05v in idle but I can't, if I try to add more value in the DVID, it make my load vcore more which I don't want.
All power saving is enabled. What I'm doing wrong?
I ever used ex58-extreme which BCLK never more than 228. I did every way to raise PCI-E to be greater than 103 but it wasn't.
Now, I use board X58A-UD7. I can raise it higher than 103 MHz and BCLK up to 242!! on water cooling
http://img138.imageshack.us/img138/1...k242resize.png
http://img715.imageshack.us/img715/6667/49ghzresize.png
Does that say 1.7 vcore??!!!?!?!?!?! Holy :banana: :eleph: :banana: :eleph: :banana: :eleph: :banana: :eleph: :banana: :eleph: Batman!
Rig is 4.2 ghz stable :)
just seems like my IOH is a bit more sensitive at lower voltages than normal- but every mobo is a bit different. running IOH at 1.14 (1.18 and above fails cold boot- warmboots fine)
warm boots are perfectly fine even past 1.3v IOH and I ran linx for 12 hours- temps low 70's on air.
I am writing to gigabyte- maybe a bios revision would allow the coldboot start check to be a bitmore lenient?
can someone speculate why warmboots are fine but cold boots fail with multiple restarts? (at higher than stock ioh voltage)
it kind of demonstrates how all hardware is tuned a bit differently depending on the parts- no two overclocks will be exactly the same...
What UD7 bios is everyone using for their 980x - and where can I download it - I should have mine this Friday.
we benched on LN2 today with the UD7 today and a 980X and ran into several issues with F6A ( dated from march) Now since it's our only I7 motherboard we can't tell if it's board or purely a so so CPU we got here... (board was modded too for the Gulfie)
anyone have concerns on the x58a-ud5 and the swiftech XT backplate in regards to the soldering that the backplate goes over?
I am concerned they may contact and short out on the backplate.
I see many of you have UD5's so i'm curious if you observed this as well.
I am using an Apogee Xt as well with no issues. I think the gulftown mod is for the tm pin to adjust for the low LN2 temps.
I've fixed the bend pins on the MB, and and now I've got six cores :D
First attempt, on air:
http://img.techpowerup.org/100316/Capture005.jpg
That's one hellova pin fix Sam :) Congrats on the new cpu, how are the temps with your Noctua NH-D14?
LO, yeah the pin-fix did some wonders. :rofl:
This is running really cool. It Idles at low 30s, which is about 10'C cooler than my W3520 @4GHz. I haven't pushed and tested the load-temp yet, but it seams to be about 15'C cooler. Using the same Noctua NHD14, push/pull. This a cool CPU. :) But I'm not pushing it yet, and can tell more later.
Got my 980 X today. Bios F6 final is not good for me. Beta bios F6x runs very well though
I haven't tried F6x, but how is it bad with 980x?
I've only tried F6 with this CPU, but my Uncore is stuck at x20, no matter what i set in BIOS it ends up like x20 in windows. I don't know if it is a BIOS-bug for Xeon DP CPUs or something else, yet. Maybe i should try F6x.