Printable View
GIGABYTE has been subject ot a lot of scrutiny lately, based on a YouTube video of a user running an Intel LCS on GIGABYTE's X79-UD3 motherboard, and the VRM's subsequently going up in smoke. This seems to be a trend (remember that video of a GTX 590 sparking and smoking?). However, GIGABYTE was quick to find the issue and correcthttp://feeds.feedburner.com/~r/NewsA...~4/WXnfBus-d1Y
More...
It is one thing to have a quick bench session but what about sustained phase/water clocks on a 24/7 basis? I'm not buying it. I don't believe a bios fix can compensate for shoddy VRMs.
they are claiming there was an error in the original bios allowing this to happen. In other words it wasn't the hardware at fault. of course it remains to be seen if they are right or just doing some CYA. but considering the overall quality of their products. i'd say give them the benefit of the doubt for little bit, and see if any other problems crop up. You know if this isn't the fix they claim there will be a ton of stuff popping up on the internet.
there is not "shoddy VRM" that was the whole point of the article. it was simply a bios bug where under specific circumstances the vrm could reach a higher than desirable temperature. This kind of situation can cause any VRM to cook. and they fixed that issue via a BIOS update. the whole point of the video was to dispell claims that the F7 bios disabled overclocking in favor of the VRM protectuion.. which as you can see from the results is completely untrue. I could go on about many previous boards from other manufacturers have had very similar issues with VRM and a bios bug causing the VRM to go on the fritz and burn, and even kill cpu's in the process but I would rather not relegate to mud slinging at this point and just more cover the bases of what is really happening and not the Recall panic some others decided to attempt to get going.Quote:
Originally Posted by PP Mguire
On top of all of this though Gigabyte still stands behind the board as well for anyone who is concerned can swap the board in no questions asked and get a replacement. So they're doing everything they could as far as I can tell to make sure customers know that they are there to support them.
By now you guys know how I overclock and If there was a bad component or an issue I will call a spade a spade but honestly I have not found an issue yet, then again I dont really put massive amount of stress on a CPU either without some sort of airflow over the VRM cooler either ;) *you might wanna watch the video of the UD3 burning again*
Few points here.
Fan blowing hard on VRM
LN2 naturally cools board around the socket including VRM
2 cores, 2 threads
Lets not forget LN2
Not on a regular basis here or fully stressed.
So like I said, I'm not buying it until I see some people doing 4+ ghz clocks on water actually stressing the whole chip without a nice fan blowing on the VRM. I can honestly say shoddy VRM because I have the same crap on my current GB board and loads of people notice throttling when it gets hot. I would expect this kind of thing from a cheap mainstream board like mine, but not by expensive stuff like this. All the bios does is make it throttle so it doesn't overheat and go kaput. Good VRMs wouldn't do that under so much stress when the board is obviously made for OC.
im not one to argue so I wont... but lets just say some of your points can be taken with a fair amount of understanding but some I can say I truly believe you may not understand...Quote:
Originally Posted by PP Mguire
the Video was made to show the BIOS overclocks... we are testing the BIOS on a UD5 now and will let you know in a day or 2 what it does on air/water....
But the initial "claim" was that the BIOS "Cripples" overclocking on the board completely.... well 5.7+GHz = not crippled. :)
BTW please feel free to discuss I have no problem just please understand I wont argue but I will be happy to discuss facts.
your board im not too familiar with tbh as I didnt do much with the P55 series honestly I spent most of that time frame on X58, but I can say with utter certainty that the P55 micro I did test from Ggiabyte ran very very well :)
http://hwbot.org/submission/911558_punx223_superpi_core_i7_860_9sec_270ms
but keep in mind since I know you will say that VRM cooled by running DICE but reality is I dont ever bench on water/air usually as its just not worth it :)
Trust me when I say I understand all about extreme overclocking. I have ran my i5 at 4.8-5ghz under custom phase, ran LN2 for a very short time (for fun) and was even considering talking to Kingpin about joining Pure once I had more experience in extreme overclocking. Life hit me in the head, and I had to drop that idea. Point is, the argument (not just mine) isn't whether or not the board can OC. That was obviously proven when people were popping VRMs, and after the bios update breaking a few records. It is whether or not it can do long term OC under less aggressive cooling measures when the heat is on (all cores going with HT) without throttle or decrease in stability.Quote:
Originally Posted by ShannonR
To put it in simple, basically the image GB is sending out to us common folk is, oh there is a problem! Ok we fixed it, see we will prove it by running 2 cores under LN2 and OCing to the max for a few minutes. K fixed!! All of us are sitting back scratching our heads saying how is that supposed to comfort the idea of me running air or water for long periods of time without extreme cooling measures to cool our board? I mean it really doesn't matter to me as I planned on buying an ASRock Extreme 9 when I upgraded, but I have people asking my opinion on this mess and the only thing I can tell them right now is wait and see.
As to my board, I knew about the weak VRM when I got it but that doesn't mean I can't run stable 4.3ghz under water. I just can't go past 220FSB without the board freakin out on me.
Not worth it to who? you personally, or the people on this site? get the board running on air or an LCS, overclock and push it so we have an answer. make it fold on all cores over clocked for a week straight and see if it pops. that will use all thread/cores to the max in a real world app.Quote:
Originally Posted by ShannonR
+1 Black!