Does the HD7970 not like as much voltage as the HD5970?

Status
Not open for further replies.

earwicker7

In Runtime
Messages
345
I'm overclocking my HD7970s, and I'm hitting a bit of a wall. I'm trying to find out what the max memory clock is using Afterburner/Kombustor; my methodology is to run it in one minute intervals, and if that works, kick it up a notch, try it for another minute, if it fails, boost the voltage and try again (I will eventually run a few 30 minute torture tests once I narrow things down, but that's down the road a bit). I can manage 1825 at 1.174V, so my next move was 1835 at 1.174V, which crashed. I boosted the voltage to 1.181V, and it crashed. Just for the heck of it, I then turned it back down to 1825 at 1.181V, and it crashed. So basically, it will run 1825 at 1.174V but not at 1.181V, which strikes me as a bit unusual, as I usually think of higher voltage as being more stable. This is all with temperatures no higher than 50 degrees. I was able to run my HD5970 at 1.3V for a couple of years with no stability problems at all, so I was assuming I could do the same on the HD7970.

Long story short... is there a possibility that this card just won't go over 1.174V, or is this just a tweaky phenomenon that may just disappear at the next highest voltage? It's two MSI R7970-2PMD3GD5 cards, and I've got a 1500W power supply, if that helps.
 
You have to realize that smaller wafers take less volts. Not every one card is the same, so one card could be unstable while the other one is causing it to work once but not work again later. Also, core voltage doesn't do anything for memory. If you are raising your voltage for memory then you are just giving the core more voltage for nothing.
 
You have to realize that smaller wafers take less volts. Not every one card is the same, so one card could be unstable while the other one is causing it to work once but not work again later. Also, core voltage doesn't do anything for memory. If you are raising your voltage for memory then you are just giving the core more voltage for nothing.

That's good to know... I guess I'll just assume that 1825 is unstable, and I just had a few lucky runs which coincided with higher voltage (they were only one minute runs, so this makes sense). I'll back it down a notch and see how that works.
 
So far I'm getting it smooth as butter (i.e., no crashing for long periods of time) at 1250/1780 @1.156V. 1780 seems to be as high as it's going to go on the memory side; the speeds I posted earlier turned out to only be stable for about five minutes. I think it's got more room on the core, just not sure if it's worth the hassle to push it any further, as two cards running at that speed seems to be quite enough to handle any games I throw at it. I'll probably change my mind and push it more tomorrow lol.
 
So far I'm getting it smooth as butter (i.e., no crashing for long periods of time) at 1250/1780 @1.156V. 1780 seems to be as high as it's going to go on the memory side; the speeds I posted earlier turned out to only be stable for about five minutes. I think it's got more room on the core, just not sure if it's worth the hassle to push it any further, as two cards running at that speed seems to be quite enough to handle any games I throw at it. I'll probably change my mind and push it more tomorrow lol.
One of those cards is enough to handle any game.
 
Status
Not open for further replies.
Back
Top Bottom