Advertisement
If you have a new account but are having problems posting or verifying your account, please email us on hello@boards.ie for help. Thanks :)
Hello all! Please ensure that you are posting a new thread or question in the appropriate forum. The Feedback forum is overwhelmed with questions that are having to be moved elsewhere. If you need help to verify your account contact hello@boards.ie

Sapphire 7950 950Mhz Edition Dual X POT. ISSUE!

Options
  • 28-08-2012 4:11pm
    #1
    Closed Accounts Posts: 6,224 ✭✭✭


    Guys a know a few of us have purchased on of these so I'd like to got hough some odd behaviour and hope it's just my stupidity.

    The card seems to come with a VDDC of 1012mV and a VID of 1.156v - I'm not actually sure what the difference is.

    In sapphire Trixx I am frequently seeing the VDDC being set to 1156 overheating the GPU even at stock/factory overclock (950Mhz).

    I am seeing issues in 2D operation - when the card is at my selected overclock - 1100Mhz/1600Mhz 1082mV VDDC I am geting very brief flashes of screen corruption. This is happening on the desktop only and is quite difficult to describe.

    I'm also getting a rather odd noise seemingly coming from the card when it draws more power - it doesn't seem to be down to any moving part.

    Any solutions / feed back welcome. Its giving me some amazing OCs but I am wondering if it heating to a point where its taking damage. NB MSI afterburner has some significant issues with the 12.8 driver and/or this card.


Comments

  • Registered Users Posts: 1,600 ✭✭✭Eboggles


    Guys a know a few of us have purchased on of these so I'd like to got hough some odd behaviour and hope it's just my stupidity.

    The card seems to come with a VDDC of 1012mV and a VID of 1.156v - I'm not actually sure what the difference is.

    In sapphire Trixx I am frequently seeing the VDDC being set to 1156 overheating the GPU even at stock/factory overclock (950Mhz).

    I am seeing issues in 2D operation - when the card is at my selected overclock - 1100Mhz/1600Mhz 1082mV VDDC I am geting very brief flashes of screen corruption. This is happening on the desktop only and is quite difficult to describe.

    I'm also getting a rather odd noise seemingly coming from the card when it draws more power - it doesn't seem to be down to any moving part.

    Any solutions / feed back welcome. Its giving me some amazing OCs but I am wondering if it heating to a point where its taking damage. NB MSI afterburner has some significant issues with the 12.8 driver and/or this card.

    Your OC might just be unstable, try bringing down the memory to 1450mhz? It's not like it makes much difference in games.

    By the way, don't use TRIXX and MSI Afterburner, as having two overclock utilites can cause trouble. I personally prefer Trixx, but it doesn't really matter.

    Also, roll back to Cats 12.7 as well, 12.8 does give me a bit of hassle when watching YouTube videos.

    What are your temps like? Really, you can pump as much voltage as you like if you can keep the temps under control. For example, if I overvolt my 6850 I can hit 1ghz but the added noise and heat isn't worth it. That said, I have the card as near to silent as I can :D.


  • Closed Accounts Posts: 6,224 ✭✭✭Procrastastudy


    Eboggles wrote: »
    Your OC might just be unstable, try bringing down the memory to 1450mhz? It's not like it makes much difference in games.

    By the way, don't use TRIXX and MSI Afterburner, as having two overclock utilites can cause trouble. I personally prefer Trixx, but it doesn't really matter.

    Also, roll back to Cats 12.7 as well, 12.8 does give me a bit of hassle when watching YouTube videos.

    What are your temps like? Really, you can pump as much voltage as you like if you can keep the temps under control. For example, if I overvolt my 6850 I can hit 1ghz but the added noise and heat isn't worth it. That said, I have the card as near to silent as I can :D.

    I even make sure catalyst is turned off tbh - multiple apps are a major issue. I just wondered if it was an issue with Trixx - there are some definite voltage issues somewhere - I think it may very well be the 12.8 driver.

    The memory has tonnes of head room - its the core that seems to be cuasing the issue. Once you start pumping in even 30 - 50mV you start to see the limits of the cooling solution on furmark (Fans at 100% increasing temps)

    I'm running in the 70s peaking in the low 80s so have started to scale back my OC and voltage. The card becomes unstable at 85C.


  • Registered Users Posts: 1,600 ✭✭✭Eboggles


    I even make sure catalyst is turned off tbh - multiple apps are a major issue. I just wondered if it was an issue with Trixx - there are some definite voltage issues somewhere - I think it may very well be the 12.8 driver.

    The memory has tonnes of head room - its the core that seems to be cuasing the issue. Once you start pumping in even 30 - 50mV you start to see the limits of the cooling solution on furmark (Fans at 100% increasing temps)

    I'm running in the 70s peaking in the low 80s so have started to scale back my OC and voltage. The card becomes unstable at 85C.
    Don't use furmark, that's not a stability test. It's to test how much punishment your card can take before crapping out. Try something like Heaven 3.0, BF3 or 3DMark11 to test for stability.

    Use Gpu-Z to measure your voltages in-game to make sure they line up with what Trixx says.


  • Registered Users Posts: 4,983 ✭✭✭Tea_Bag


    didnt AMD set it up so that your cards get crippled when it detects furmark is running anyway? i know nvidia has done it on the 600 series.


  • Closed Accounts Posts: 6,224 ✭✭✭Procrastastudy


    What does Board Power Li do? I know its something to do with thermal properties...


  • Advertisement
  • Registered Users Posts: 1,600 ✭✭✭Eboggles


    What does Board Power Li do? I know its something to do with thermal properties...

    I'm pretty sure it's a power limit, you should have it at +20% as it gives you much more to work with OC wise. It could be the source of your problem actually.


  • Registered Users Posts: 1,600 ✭✭✭Eboggles


    Tea_Bag wrote: »
    didnt AMD set it up so that your cards get crippled when it detects furmark is running anyway? i know nvidia has done it on the 600 series.

    Yup, I'm pretty sure they did. The only reason anyone should use Furmark would be if they wanted to warm up their room in the Irish Winters :D.


  • Closed Accounts Posts: 6,224 ✭✭✭Procrastastudy


    Eboggles wrote: »
    I'm pretty sure it's a power limit, you should have it at +20% as it gives you much more to work with OC wise. It could be the source of your problem actually.

    Set to +20% already. There is something funky going on with this card.

    I've mianly been using Heaven to test - it doesn't really have any issues - just didn't put as much thermal load on the GPU is all.


  • Registered Users Posts: 1,600 ✭✭✭Eboggles


    Set to +20% already. There is something funky going on with this card.

    I've mianly been using Heaven to test - it doesn't really have any issues - just didn't put as much thermal load on the GPU is all.

    The temperatures you're getting on heaven are probably going to be the highest you'll even get while gaming, what are those temperatures by the way?


  • Closed Accounts Posts: 6,224 ✭✭✭Procrastastudy


    Eboggles wrote: »
    The temperatures you're getting on heaven are probably going to be the highest you'll even get while gaming, what are those temperatures by the way?

    Actually BF3 and GW2 are pushing it higher funnily enough. at 1100/1600 1082mV I'm seeing 81C max at abour 80% fan.

    over 100A VDDC current - thats not right is it?

    EDIT: I *THINK* the brief flash on 2D mode is a very brief ramp up from low power mode with the voltage not responding quick enough.


  • Advertisement
  • Registered Users Posts: 1,600 ✭✭✭Eboggles


    Actually BF3 and GW2 are pushing it higher funnily enough. at 1100/1600 1082mV I'm seeing 81C max at abour 80% fan.

    over 100A VDDC current - thats not right is it?

    EDIT: I *THINK* the brief flash on 2D mode is a very brief ramp up from low power mode with the voltage not responding quick enough.

    If you're really unsure just revert to stock clocks, and try it for a day or two. If you're not getting the problems, it's an unstable OC.

    Actually, you might need some more voltage :D. After looking up some logs it seems some people need 1.2v to get 1100 on the core.

    What's your airflow like inside the case?


  • Closed Accounts Posts: 6,224 ✭✭✭Procrastastudy


    Eboggles wrote: »
    If you're really unsure just revert to stock clocks, and try it for a day or two. If you're not getting the problems, it's an unstable OC.

    Actually, you might need some more voltage :D. After looking up some logs it seems some people need 1.2v to get 1100 on the core.

    What's your airflow like inside the case?

    @1100 I'm seeing it hit 1030 ish peaks at 1051 as low as 1018 - seems its the VRM thats heating up the GPU got to 72C max where as VRM1 is showing 79C.

    I'm not sure how it can be an unstable OC it runs fine in games - its just in 2D - and I think it's just when the card ramps up. I've disabled ULPS - going to see if that helps.

    Thanks for the help EB its appreciated - any further thoughts?

    EDIT 600T with a H100 sucking air out and a 200mm + 120mm intake. I also have an additional 120mm exhuast and the mesh on the side panel.


  • Registered Users Posts: 1,600 ✭✭✭Eboggles


    @1100 I'm seeing it hit 1030 ish peaks at 1051 as low as 1018 - seems its the VRM thats heating up the GPU got to 72C max where as VRM1 is showing 79C.

    I'm not sure how it can be an unstable OC it runs fine in games - its just in 2D - and I think it's just when the card ramps up. I've disabled ULPS - going to see if that helps.

    Thanks for the help EB its appreciated - any further thoughts?

    EDIT 600T with a H100 sucking air out and a 200mm + 120mm intake. I also have an additional 120mm exhuast and the mesh on the side panel.

    Those coolers should be performing better than they are, and with the 600t being a relatively good case along with some fans you should be getting better temps.

    The VRM can go up to 105c safely, so that's not a problem.


  • Closed Accounts Posts: 6,224 ✭✭✭Procrastastudy


    The fans are getting quite loud at high RPMs I wonder if they are banjaxed.


  • Closed Accounts Posts: 6,224 ✭✭✭Procrastastudy


    I'm wondering if I have a faulty VRM - VRM 1 is showing a much higher temp than everything else. It might also explain the odd noise.


  • Registered Users Posts: 1,600 ✭✭✭Eboggles


    The fans are getting quite loud at high RPMs I wonder if they are banjaxed.

    I'd say they're getting louder because they're spinning faster :pac:

    EDIT:VRM temps are normally higher, but they cope better with higher temperatures so it doesn't matter.


  • Closed Accounts Posts: 6,224 ✭✭✭Procrastastudy


    Eboggles wrote: »
    I'd say they're getting louder because they're spinning faster :pac:

    EDIT:VRM temps are normally higher, but they cope better with higher temperatures so it doesn't matter.

    I'm gonna keep playing with it - I think some of the issues are down to the driver. I dunno about temps - it seems that once toy start pushing into 1100-1200 teritory the GPU really starts to sweat. You can run at 1156mV at 950Mhz alright but try and put that through it at 1200Mhz and its going to slap you back.

    That said even 1000Mhz is a 20% OC for that chip so I shouldn't complain.

    My only query is - given its the same GPU as a 7970 shouldn't it be able to take the same voltage and clocks?


  • Registered Users Posts: 4,983 ✭✭✭Tea_Bag


    wont be as well binned as a 7970.

    kinda sounds like a driver issue to me, or youve managed to physically damage the chip by pushing it too far. have you tried the new beta drivers?


  • Closed Accounts Posts: 6,224 ✭✭✭Procrastastudy


    Tea_Bag wrote: »
    wont be as well binned as a 7970.

    kinda sounds like a driver issue to me, or youve managed to physically damage the chip by pushing it too far. have you tried the new beta drivers?

    Would pushing it to far cause run away heat issues - I think thats what I did to one of my 4870s.

    EDIT regarding the flashing on screen in 2D - I'm pretty sure that was a power state thing.


Advertisement