GA-990FXA-UD5 issues (need quick answers)

[autotech]

My thoughts are when they test boards they do it all at default setting. Dont you remember all the overclock warnings   Do at Your Own risk.

[zuban]

the voltages are wrong at default.

[gareth170]

My thoughts are when they test boards they do it all at default setting. Dont you remember all the overclock warnings   Do at Your Own risk.

even if they test boards at default/stock settings they would of seen the voltage is way off and drops badly.

i had a system crash when the system was under load due to the bad vcore drop under load, and this was at stock settings .

i don't  mind the 0.044v that it adds on , but the bad vcore drop under load i do mind.... a 0.060/0.100v drop isn't right..... i accept if it's 0.001/0.003 drop.

[zuban]

thats what gets me the most with this, it isnt a problem with overclocking, its a really basic issue.... I can not help but look at it and think they have tried to cover this issue by upping the voltage a certain amount above what it is set in the bios in an attempt to cover off the voltage drop, i'm not columbo, but unless this is a bad batch of parts that looks very much like what has happened, and if true its quite frankly shocking, conjecture or simply a theory we wont know until gigabyte confirm what the problem is .

[jolphil]

I have read this thread and my  first point is, that I am not in any way connected with Gigabyte and only own a GA870A-UD3 board with a 1090t CPU.I think it is totally unrealistic to think that the core voltage will vary from .001 to .003 volts or 1 to 3 millivolts even if "Cool and Quiet' is disengaged"..From my own tests I determined the CPU at idle will dissipate somewhere in the neighbourhood of 60 watts and at full throttle around 120 watts..
If you assume some average core voltage say 1.45 volts, that translates to about 41 amperes to 82 amperes at wide open..This core voltage comes from the +12 volt line where it is created by an array of power FET's to each turn on and off to average,filter and regulate the lower core voltage..Each FET has a Rds (FET on resistance) and also consider the copper path from the FET's to the CPU..Also consider if the power supply 12 volt source drops as load increases..Not ever mentioned is where the voltage sensors are placed and how accurate they are to begin with.. these factors plus more not mentioned can determine the variance of core voltage..We can only assume that Gigabyte as well as Asus and others test their boards for the worst case conditions that will allow stable operation..
With todays low voltage high current CPU's I fully expect there will me more variations then yesteryear..
I am not in any way suggesting you do not have a bad Motherboard but the expectation of that very small variance is in my opinion unreasonable..Granted you should not get blue screens at wide open but there are other factors to also consider.Temperature of both CPU as well as chipset may also play a part..
I hope you iron out your difficulties,be patient and methodical and most of all Goodluck..
jolphil

[zuban]

at stock my cpu which is supposed to be a 1.4v cpu is set to 1.52v, it drops to 1.46, then 1.44, then 1.47 and is quite frankly all over the place if it doesnt crash first, it is never stable, it is always a long way off whatever is set in the bios, given i had a gigabyte 880gm-ud2h board before this that had none of these issues and could even clock my 1090t higher than the ud5 i think there is without a doubt a problem with this board. If it was just me having these issues then that would seem ok i am just unlucky but i have seen so many other people have this exact issue.


thankfully my retailer will take this board back and have offered me a refund as they agree it is faulty, if i dont hear from gigabyte on monday it will be going back.

[gareth170]

I have read this thread and my  first point is, that I am not in any way connected with Gigabyte and only own a GA870A-UD3 board with a 1090t CPU.I think it is totally unrealistic to think that the core voltage will vary from .001 to .003 volts or 1 to 3 millivolts even if "Cool and Quiet' is disengaged"..From my own tests I determined the CPU at idle will dissipate somewhere in the neighbourhood of 60 watts and at full throttle around 120 watts..
If you assume some average core voltage say 1.45 volts, that translates to about 41 amperes to 82 amperes at wide open..This core voltage comes from the +12 volt line where it is created by an array of power FET's to each turn on and off to average,filter and regulate the lower core voltage..Each FET has a Rds (FET on resistance) and also consider the copper path from the FET's to the CPU..Also consider if the power supply 12 volt source drops as load increases..Not ever mentioned is where the voltage sensors are placed and how accurate they are to begin with.. these factors plus more not mentioned can determine the variance of core voltage..We can only assume that Gigabyte as well as Asus and others test their boards for the worst case conditions that will allow stable operation..
With todays low voltage high current CPU's I fully expect there will me more variations then yesteryear..
I am not in any way suggesting you do not have a bad Motherboard but the expectation of that very small variance is in my opinion unreasonable..Granted you should not get blue screens at wide open but there are other factors to also consider.Temperature of both CPU as well as chipset may also play a part..
I hope you iron out your difficulties,be patient and methodical and most of all Goodluck..
jolphil

it seem some here like to blame everything else but the motherboard.

i know alot about hardware. has for the psu i have a Corsair AX1200 1200w and the +12v line doesn't hardly drop under load so it's not that,, also my cpu temp is around 44c under load so it's not that.

it isn't unrealistic to say that the vcore shouldn't drop more than 0.001/0.003.. my other motherboard gigabyte 890fxa-ud5 has NO vcore drop under load, not even 0.001v, so theres NO reason why this motherboard shouldn't be like that.

to be honest, all i want is for someone from gigabyte to come on this forum and say "that they working on the vcore issues via bios ,  but if it turns out to be a design flaw then they'll replace the motherboard with the next rev for free.."

simple as that

[jolphil]

it seem some here like to blame everything else but the motherboard.

If you read carefully my post I did not suggest that it could not be the motherboard and in fact I said....

I am not in any way suggesting you do not have a bad Motherboard but the expectation of that very small variance is in my opinion unreasonable.

My point is and was that most motherboards makers  do vary the V core voltage with CPU load .. Usually they increase the Vcore with load
depending on the CPU chip as well as other factors..So it could very well be that your Motherboard is defective because your Vcore drops under load..It also could be that the CPU may draw more current than the supply can deliver..AMD CPU's will throttle back if the temperature gets too high ..These are just possibilities not Fact..Again you may have a defective board..Who Knows..
There is one way to find out and that is to RMA the board ..

Here is just one URL that shows other users complaining about varing Vcore..It is just one example of my point..There are four(4) pages so please read all four..
http://www.overclock.net/amd-cpus/751265-1055t-vcore-goes-up-load-turbocore.html
Again good luck and I hope you find out whats wrong
jolphil

[absic]

As I stated previously, I do believe this issue is BIOS and not hardware related.

The VCore voltage drop is being shown across several different PC configurations and in testing on my own set up I have encountered a drop in VCore voltage when stressing the CPU which was not shown on the GA-890FXA-UD5.

The primary reason I believe the BIOS is at fault is because I have been testing this board for two weeks and have encountered other issues with the BIOS which I can best describe as "glitchy". The hardware I am using has been tested (over the past 10 days) on a series of Gigabyte motherboards and they have all worked perfectly. It is only on this board that I have encountered any problems.

Gigabyte are obviously trying to sort things out, hence them sending me the F5a BETA BIOS but unfortunately the issues are still there. Gathering information, testing, writing a new BIOS and all of the steps involved in resolving problems such as this do not happen over night. If you are not happy then, by all means swap your board for another. Or, you could do as I am doing, gather as much accurate information along with screenshots and pass this back to Gigabyte to help them with helping us. The more accurate information they receive the easier it will be for them to diagnose the cause and come up with a solution.

Hopefully, it is not a hardware issue but if it is, then like the recent problems with Intel's chipsets, I am confident that Gigabyte will do the right thing and replace the boards.

[Night Gyr]

Good morning Ladies & Gentlemen!

Unfortunately, I'm inclined to perceive the sub-par performance of the power delivery subsystem of the recent Gigabyte motherboards (i.e. those using the new Dr. MOS tech) as a trend.

There was an in-depth review/comparison of the GB P67 UD7 board and the ASUS ROG P67 board. Now even though the UD7 has 100+ phases, Dr. MOS and an all new Intersil controller, its power system performance was much worse than that of the ROG board. To be exact, it demonstrated random oscillations and fluctuations of VCore. ASUS board, which uses a relatively simple digital power subsystem, performed much better.

It's very sad to see posts like this, because they make me believe that the intensive promotion of the new Dr. MOS & friends tech is nothing but marketing hype, and that the famous Gigabyte quality is slowly becoming a thing of the past, simply a legend.

Coming back on topic, absic, could you please explain why would BIOS affect the CPU power delivery? IIRC the controller has its own BIOS of sorts, which can not be easily replaced or upgraded (if possible at all.) And a BIOS update doesn't seem to upgrade the controller microcode I think.

Regards,
--Night Gyr

[gareth170]

@absic  , but surely if it was a bios issue why would the 990fxa-ud7 have the same issue, but the 990fxa-ud3 seens that it doesn't have the issue.

surely the ud7 doesn't use the same bios has the ud5..

[absic]

surely the ud7 doesn't use the same bios has the ud5..

I would imagine the base BIOS for all four of the 990FX  boards is pretty similar as the basic hardware and CPU options are the same. All these boards use the same Northbridge and Southbridge chips as well as LAN, Audio &  USB3 chips. Yes, the configuration is different but the underlying hardware is the same.

@absic  , but surely if it was a bios issue why would the 990fxa-ud7 have the same issue, but the 990fxa-ud3 seens that it doesn't have the issue.

With this question you have to consider the fact that not too many of these boards are out there at the moment and, of those that are, not everyone will be examining them with as much scrutiny as we are.There are probably plenty of users blissfully unaware of the VCore Voltage  and other BIOS problems who are running their systems quite happily in ignorance. The fact that issues are not being reported here, or on other forums, doesn't mean there aren't any. I first reported my concerns to Gigabyte (via runn3R - The forum moderator) on the 12/6/2012, a clear 12 days before you started this thread. At the time I was unaware of any other user having problems despite searching the internet and other forums and thought it might just be down to my own choice of  hardware. Only when you, along with other's posted was it apparent the problem was not only on my system but on others too.

[gareth170]

ok..

i thank im keeping the board because it's a big hassle changing motherboards, has it takes over 3 hours to install a motherboard due to a woman doing it for me (being  my hands)...

but theres no way im letting this go. i'll give gigabyte a chance with a new bios and if it doesn't  fix it i will want gigabyte to replace the board with ether a new rev, or a working rev 1.0 board which doesn't drop the vcore under load..

i'm being fair..

because has it is now it's not 100% stable at stock under fully load due to the vcore badly droping under load.

[Dark Mantis]

Ok I know you are good absic but I never realised that's how you got your answers. Don't you think that time travel is cheating just a little bit ?
 

[zuban]

i think i'm just going to send the board back and go for another, dont see the point in testing this board for gigabyte and there is no way they didnt know about this problem before release and the more i think about it the more disgusted i am about it. I feel part of gareth170's frustration about this though,i beleive he knows what he is talking about as he helped me understand this problem in the first place, its a an inconvenience to change mobo at best, as far as i'm concerned this should not be happening.