D0 I920 and Gigabyte EX58 UD5 issues.

I have already shipped the CPU off Plec.. I wish that I had decided to try this before I pulled the CPU out of the mobo...

It probably wouldn’t have helped, Dave, you made the right decision – it was time to take a step back and let someone else take the strain. It’s just a computer – **** it! – it’s not worth the stress it’s clearly put you under.

But surely if these pins weren't touching the processor would fail to work completely?

Or would something like this cause the type of problems that I am experiencing??


The way these new chips are designed to fit into the mobo I can imagine your paranoia as they are not as satisfying to fit as the old multi long pinned efforts.

I’m afraid I can’t answer that - other than I would have thought it would either: not post, or the cpu would be able to identify that it had a connection problem if it got into a bootable environment - beeps or BIOS message etc...

But it is plausible that it may not have been connecting properly – as to what effects that would cause I couldn’t answer definitively.

You could try a post on the CPU forum – but you may just add to your stress levels if you get a reply that contradicts your decision to rma the cpu. Perhaps it’s a question best saved if your cpu gets returned?

I would rely on the premises that: 'ignorance is bliss’ and ‘see no evil - hear no evil’ :- until you get a response form OC’s. That way you won’t run the risk of being tormented by conjecture while your chip is away being tested.

If the CPU does get returned – it would be worth refitting/posting a thread and testing the theory out – but for now I would give yourself a complete break for a few days.
 
Last edited:
Yeah, you are right man, I am going to just forget about my rig until I hear back from the people here whether the CPU is faulty first before I take any other steps...

I will keep you updated on this thread Plec...
 
Hey Plec, Still no word on my processor as yet, it has been with the OcUK guys since last Saturday, and I called them today to see if there was any news on it, and the guy said that it is still being tested and that it usually only takes them a couple of days to test goods...

Hopefully I will get news sometime in the beginning of next week...

I will let you know as soon as I find out myself..
 
So the gun is still holstered, for now - good to know. ;)

Hopefully, all this perlonged testing is a good sign and the chip will report back faulty and you can finally get to grips with this build.

Roll on monday...
 
Heres hoping man!!

I am having to use my Fiancees PC atm, and I hate it... I just want my i7 rig working as it should...

I'm going to order myself a Titan Fenrir HSF for my CPU either this week coming or the following week so that I can start having fun overclocking my rig... that is if everything works as it should when I get a replacement processor sent out to me!!
 
Well, thats my CPU being returned to me with no faults found whatsoever...

Said they ran Prime95 and OCCT for several hours with no problem... Needless to say I am ******* raging about this, because I have managed to run both OCCT and Prime for hours on end with the RAM @ 1066MHz...

I really dont know what else I can do here... I have spent £1700 on a rig that doesn't want to work for me...
 
B*ll*x bum!

Sorry to hear that Dave - the CPU was a long shot – but your reasoning behind returning it made sense so don’t feel too bad.

Said they ran Prime95 and OCCT for several hours with no problem... Needless to say I am ******* raging about this, because I have managed to run both OCCT and Prime for hours on end with the RAM @ 1066MHz...

Did they mention what settings they tested it at – I would be surprised if they ran it at 1066MHz as it’s not classed as clocked unless it’s over 1333MHz.?

I really don’t know what else I can do here... I have spent £1700 on a rig that doesn't want to work for me...

If they did get it stable at 1333MHz then you’ve run out of components to rma/test other than the motherboard?

Unless, you have one of those very obscure problems where an innocent looking cable is causing all your grief (have you tried changing all the internal sata cables and the kettle lead at the back?) the motherboard has to be the cause of the problem by process of, a very stressful and lengthy, elimination.

I know the cable test seems an inane test – but you’ve had such bad luck with this and worked really hard trying to resolve it – it would be a real b*st*rd if you returned the mobo and it turned out you had one of those one in a million cable faults.

If your motherboard got returned back to you as fault free – I have this vision of you driving up the wrong way of the M25 humming the Windows log in… :/
 
Well, I am even more p***ed off today, as I have just recieved my CPU back this morning and inside the package was a couple of pics of screen shots of OcUK technicians "thourough testing"...

All they did was ran Small FFT's for 2 hours 44 minutes, and 32bit Linpack in OCCT for 2 hours 55 minutes!!!

I made it very clear to them about the problems that I was having, and also how I could run Small FFT's for 24 hours without any issues, and that it was Prime Blend and Large inplace FFT's that caused my problems...

I personally don't class this as thorough testing.. I mean, I did all this myself, and also they had my CPU for almost 2 weeks only to test it for a mere 6 hours on the two tests that I made very clear that I could get it to pass on myself?!?!?!

This kind of defeats the purpose really... I mean if I was the one testing goods that people think are faulty, I would run the tests that the person had problems with, not the ones that they can run without any issues... this has proved absolutely nothing!!!

And I called them up today only to get told by the guy that I would have to send it back yet again for further testing??!??!?

Why cant your technicians do proper testing in the first place instead of messing people around like this???

I personally feel like you are taking the pi$$ out of me here, by not running the tests that I stated were causing the problems for me!!!

Anyway, I am about to rip my rig apart and build it up outside the case with only the essentials hooked up and see how I get on with that...

I thought that you guys would have done better tests that this, especially when I sent about 3 seperate webnotes about the problems that I was having, explaining in detail as to all the problems that I was experiencing with my system...

This will be the last thing I ever purchase from OcUK, I will take my custom elsewhere in future...

Thanks for nothing!
 
I can understand your frustration and very apparent anger – but, if you were to send it back again, you’re almost guaranteed a thorough test this time. Plus you’ll probably get it tested for free as they will have your previous web notes on record so can validate your claims. I realise this is scant compensation for your time and stress – but it may make the whole process a little easier to swallow - maybe… (I’m not defending what they did – I’m just trying to placate you a little before your next post ;))

Anyway, I am about to rip my rig apart and build it up outside the case with only the essentials hooked up and see how I get on with that...


The skeleton setup outside the rig is definitely the way to go forward, especially if you’re undecided about which piece of hardware to rma.

If you have alternate SATA cables and kettle lead I would experiment with these too as you’ll be stripping the rig down anyway – may as well rule those out too at the same time.

I keep saying it – but I mean it – good luck with it - and try to keep things in perspective while you’re imagining smashing it all up into a myriad of different sized pieces.

As I mentioned before, I really don’t want to see you on the BBC’s live feed driving up the wrong way of the M25 humming the Windows log in… ;)
 
LOL@ your comment!!! :D

Well, I am just away to go and setup my PC outside the case, and test my HDD's with different SATA cables to see if it is that which is causing the problem...

About the PSU "kettle" cable Plec, the Be Quiet! PSU doesn't use the normal power cable, it has this really chunky looking lead with a plug that I haven't ever seen before... it looks similar to a kettle lead, but it is chunkier and the pins are different from a kettle lead to....

Although this rig is really stressing me out, I know that I will get this problem sorted one way or another...

I am just trying to stay calm about it all and keep a clear head, otherwise I wont be able to focus when I get round to setting it up outside the case and then I will probably end up making things worse!!!

I will say it again, thanks so much for being here with me through this Plec, you have been really patient, and a lot of help to..

I will keep you updated man...

And dont worry, it hasn't got to the stage of me completely losing it... yet!!!

And I dont drive neither, so you're safe enough!!!! :D
 
Last edited:
Forgot to say as well, but when I opened the box to check my processor, it took out the piece of foam stuff to find that the CPU was still covered in TIM...

Maybe this isn't that big of a deal, but I would have preferred my items to be returned to me in the condition they were when I shipped them back in the first place... not covered in AS5!!!!
 
I have same hardware, here is some tips:

First: Clear your CMOS from the bottom at the back of your motherboard so we can start from the beginning
Second: Enter the bios and "Load optimzed default", save and exit
Thrid: Enter the bios and "Enable XMP Profile1", save and exit
Fourth: Enter the bios, set your memory timmings manually 8-8-8-24-2T save and exit
Fifth: Enter the bios, set your sub memory timmings manually set tRFC to 88 and leave other timmings at Auto. Save and exit
Sixth: Set your QPI/VTT voltage to 1.355v and leave all other voltage settings @ Auto
Seventh: Enter the bios then Advanced CPU Features, and disable 1-Intel(R) Turbo Boost Tech 2-CPU Enhanced Halt (C1E) 3-C3/C6/C7 State Support 4-CPU EIST Function. Save and exit

I had the IRQL_Not_Less_or_Equal BSOD, it comes when I set unstable Uncore frequency and when I use low QPI/VTT voltage
 
OK, I have spent the entire day tinkering with my rig, I decided not to rip the mobo out of the case, and instead I just disconnected all front panel switches/USB/Firewire cables, and also disconnected all case fans, and just left my boot drive connected and began testing at stock CPU frequency and the RAM running @ 1600MHz with all timings/advanced timings set manually, and Vcore, QPI/VTT, and DRAM voltages set manually to..

I booted into Vista and began a Blend test in Prime95, and it stopped running after 1 hour because of some errors, my system didn't BSOD, freeze or reset itself, Prime just stopped running...

I tried loosening up RAM timings as well, but the same thing kept happening...

So I decided to lower the CPU multi to 19x and raise Bclk to 160, I also dropped the DRAM voltage from 1.655v to 1.58v and QPI/VTT @ 1.395v, giving me a 3GHz overclock with my RAM running @ 1600MHz..

Booted into Windows and began Prime Blend Test again, and it has now been running for just over 5 hours without any issues...

I don't know exactly why it seems to be stable now, but I would have thought with the DRAM voltage being set lower than what it supposed to be I would have experienced even more instability, but this doesn't seem to be the case at all...

It will be left running this Blend Test overnight, and I will then spend the whole weekend testing my rig with these current settings to see whether this isn't just a lucky run or not, but if it starts misbehaving once more over this weekend then it will be getting torn apart again and the mobo will be getting shipped back to the retailer for proper testing...

@boldenc, thanks for the tips, but I am interested to know if there really is any need to save and exit the BIOS after making each adjustment???

I personally would just change all those settings at once, and save and exit after making all my adjustments... What difference would it make by making one change, saving and exiting, then re-entering the BIOS making another change, saving and exiting and so on??

Am I missing something here?

Anyway, hopefully I have finally found stable settings for my system.. I really want to get myself a new HSF ASAP, because I would love to have my rig running @ 4GHz.. my chip can reach 4GHz no problem at all with only 1.200v, but my Zalman cooler just isn't able to keep the chip cool at this frequency...

I might manage to buy a Titan Fenrir next week hopefully!!
 
Well, the problems are still present.. although they dont seem to be happening as often as they were before...

I am just going to leave it as is for the time being until I can get a new RMA number for the mobo to have it shipped off for proper testing...

When I decide to do my next big upgrade again, I wont be purchasing items from different retailers, I will make sure that they all come from the same source... infact I will probably just buy a mobo/ram/cpu bundle when I do another build, because I dont ever want to experience anything like this again!!!
 
Balls – I was hoping a combination of what was suggested by, boldenc, and your own blend of additional settings may have got you some results.

Dave, I don’t want to p*** you off more than you already are with this rig (if that is possible)– but I fear the following might:

If you can get the system stable at stock speeds with the memory set at 1333MHz, or 1600MHz in your case, (as technically anything over 1333 is classed as an OC – daft I know) - there is a slim chance that the e-tailor may test the motherboard and class it fit for purpose – even if you don’t.

Don’t misinterpret the above, I’m 100% on your side here, as you have bought a premium motherboard which advertises itself as a clocker’s motherboard – but the trouble is its results rely on other components to achieve this and the e-tailor could very well use this against you if it passes a default setup test. :/

Before you implode - read on...

With the above in mind, before you rip it all apart again – run the rig on default (if you haven’t done so a million times already). Using recommended manufacturing voltages and settings for all components. Set the memory at 1333MHz – I know this is galling but it’s how the e-tailor will, probably, test it and then check and hope that the stability is just as flaky.

Hopefully this will prove to be the case – and, if it does, make detailed and comprehensive notes of this in your rma including details of the other components that you’ve had tested and replaced in your bid to get it stable.

With all the testing and replaced equipment; if it fails on stock – you will have a very compelling argument for a straight swap regardless of their results.

Basically – if it fails a default setup with the new memory, new psu and the cpu passed as fit’ish by OC combined with all your troubleshooting – return the mobo with the expectation that you expect it to be replaced.

Especially, when considering that the combination of your components should technically clock to near 4GHz with very little effort at all.

Seems weird to be saying this – but I hope it fails the stock test…
 
Well, the mobo has definitely had it.. for some strange reason the 1st DIMM channel has packed in on me..

I noticed today when I restarting my system after Prime crashed, I pressed TAB so that I could see the POST screen instead of the Gigabyte Splash Screen, and it was only seeing 4GB of RAM...

At first I thought that one of my modules was dead, but I removed the Dominators and installed my XMS3 3GB set, and the same thing happened, it would only recognize 2GB of my RAM...

I then shutdown my rig and pulled the power cable out, and removed 1 stick from the 3rd memory channel, and started up again, but it would only recognize 1 stick of RAM..

So I shutdown yet again, and put the stick back into the 3rd channel and pulled the stick that was in channel 1 out of the system and booted up again and this time it recognized both sticks... I then tested each stick of RAM out in the 2nd and 3rd white slots on my mobo and they all work fine....

Also I tried having 1 stick of RAM in the blue slot on the first channel and the other 2 modules in the white slots on the other 2 channels and again it wouldn't see the stick installed on channel 1, but when I moved the stick from the blue slot in channel 1 to the blue slot on channel 2 it would pick up all 6gb but only in dual channel mode obviously...

It is strange though, because in the memory tab on CPU-Z it says that there is 6GB (or 3GB with my other set of RAM) installed on my system, and also in the basic information screen in Vista it also says there is 6GB installed, but on the POST screen and also in the Performance tab in Task Manager they both only see 4GB with my Dominators and 2GB with my XMS3 set...

Anyway, I am just about to go and try booting it up with only 1 module installed in channel 1 to see what happens..

What would cause 1 of my DIMM channels to pack in like this??

I am going to give the e-tailers a call and tell them about this problem so that I can get a replacement mobo sent out to me... what a PITA!!!
 
Last edited:
Oh, I forgot to say Plec, I tried testing with optimized defaults in the BIOS, and with the RAM running @ 1333MHz and 1600MHz, and sometimes it will run for a few hours before it BSOD's, other times it will just lockup my system forcing me to restart, and then the next time it will reset itself...

Pretty much the exact same issues I have been experiencing all along TBH...

I know this might sound weird, but I am actually happy that I have lost a DIMM channel on this mobo, because at least now I know that there is an actual hardware problem with the mobo and it isn't down to bad BIOS settings that is causing instability...

I think this mobo has been flaky from the very beginning, and this has been what is causing all these problems I have been having with my system...

Anyway, I will let this be a lesson to myself... never again will I purchase my components from different e-tailers, when I start my Fiancee's i7 build, I will probably just buy the overclocked 4GHz bundle from here tbh...

Also, I must apologize for my comments yesterday, I was just really angry, and I shouldn't be complaining about the technician leaving AS5 on my CPU's IHS... I was just looking for a way to vent my anger, and I am really sorry for being an A$$!!!

At least it saved me using some of my own AS5!!! :D
 
Back
Top Bottom