NVIDIA Green Light program requires board partners to validate their designs
Click here to post a comment for NVIDIA Green Light program requires board partners to validate their designs on our message forum
Mufflore
From the bolded
The point of them making a statement is that they have influenced MSI through the warranty.
You'd think they would be good at logic :P
DarkKnightDude
I remember there was a rumor going around that the 680/670 series that performance degraded over time through overvolting, maybe this is related.
I don't see why EVGA or MSI would put money and effort into creating the Classy or Lightning and then disabling their over volting features later.
Pill Monster
Oh Snap - I just posted a link to that article in another thread. π
You think you'd still say that if your card died 6mths from now?
Personally I like a little bit of security.
Pill Monster
Oh, almost forgot - seems appropriate;
http://gentoochan.org/g/src/1340775015574.png
Veteran
I would have liked a voltage mod for the 690 as it would then justify watercooling it for myself but atm using water and only a 30mhz increase on average over air just isnt worth the time and hassle even though i already have pre-purchased all my custom water equipment,guess ill save it for something better down the road.
buddybd
I don't see the fault in this really. If other companies want to mod the heck out of their cards then why should nvidia have to pay for their mistakes? If they are absolutely confident that their specs won't blow up the hardware, then they can safely provide the warranty themselves.
But clearly they don't trust their cards as much as they would like to. So yea, it is all about the money. Both on Nvidia and AIC's part.
BLEH!
Isn't this kinda thing fixable at the BIOS level, as in, if someone writes a custom BIOS for the nVidia cards?
PhazeDelta1
Probably. Someone on the EVGA Fourm said he was working on a bios editor for the 600 series cards.
PhazeDelta1
---TK---
almost seems like a ploy for free advertising by nvidia, as soon as the bios editor is confirmed working people are going to buy these cards out of spite just to overvolt them.
rflair
Moderator
Neo Cyrus
Sounds like nVidia are not confident in their products. I've had this horrible excuse for a video card overvolted to hell and back and I've had it since the start of 2011 (IIRC), it shows no signs of degradation.
Wow, I never even thought of how long I've had it, I've never had a high end video card hang on this long. Usually I have to replace them after a year or suffer crappy performance in the latest games. That just shows how horribly progress has slowed down and I blame it all on consoles and their crappy 2005 technology.
Lane
Funny enough, processors, GPU, etc are tested and made for work with higher voltages of what is set ... the stock voltage is part of how they bin the card between TDP respect, and stability...
I tend to ask me if Nvidia have not been too much short on some specification of the PWM system .
But this should not inclue high end overclocking card with own brand pcb as the Evga, the lighting etc..
AMD cards have a limit max on overvoltage too.. for the 7970 it is fixed by AMD at 1.381V for reference cards ( who is really high anyway ).. Hence why this is the max you can obtain with MSIAB or Trixxx .. It is free to the AIB if they allow in the bios an overvoltage higher to get component for do it. ( Lightning and other gpu with Power controller different of the reference one ).
But lets be serious, many years ago, i remember have kill many cards, including 2x ATI 9700Pro Maya edition and 1x ATI 9800XT, Nvidia 6600GT, cause at this time the cards had absolutely no protection, software or hardware. today the cards are really more safe about abnormal condition.. OCP, temperature protection, tdp limit etc etc ....
Speaking about RMA for Nvidia look for me a bit strange. they are binned as they respect the tdp and clock speed at a voltage x... But they are designed for be able to take a lot more of voltage. And with the last feature as TDP limit, control + Boost clock, thoses type of safety real time control, work even deeper of the simple " protection":
If i take the exemple of AMD, If you reach a high temp, or have set too much voltage, on a certain point, the card will by itself decrease the voltage and clockspeed dynamically for adapt them to different factor. whatever you have set. for protect the "card".. This is the type of protection got the cards today and we had not before.
The problem with the MSI 660 Power edition card, for what i have see, is they use an additional piece on the voltage controller for push 9.3V on the Richtek who work with 5V recommended and 7V as maximum .. I dont know if the 2.3V more can or will cause damage in a long term use.
For the EVGA; the problem is ofc different.
Mufflore
My take on it is this:
People worked out how to make gfx cards run faster.
This generated a market that gained competition, mfrs provided the tools to do the job because customers demanded it.
NVidia set the limits for clocking based on what wont harm the cards but tried to give us the max possible, so they are pushing the boundaries.
Sometimes the limits are set a bit too high, but it takes time to find out some problems.
Thus to remain safe, the limit has to be lowered if a problem appears.
I dont see there is a need to lay blame, its just one of those things.
They have been nice enough to us, sometimes we cant have it all.
sykozis
I'm kinda with Phase on this one. If I want to push 12V through my GPU....where does NVidia have the right to tell me I can't?? I own the card. They forfeited all rights to it the day I bought it.
rflair
Moderator
Pill Monster
I'd be interested to know how the the supposedly reduced number of RMA's will stack up against expected reduced sales from the cards being castrated....
This can only be good for AMD.
Denial
Pill Monster
sykozis