4 Replies Latest reply on Jun 28, 2016 5:23 AM by alandolby

    R9 270X: 0.950V instead of 0.875V in 2D Mode

    alandolby

      Hi guys and gals,

       

      I'm having a minor issue with my Sapphire R9 270X Dual-X 2GB OC card:

      While in 2D mode the VDD voltage is higher that is should be, at 0.950V while Clock and Memory are at the correct 350MHz and 150MHz respectively.

      Voltage should be at 0.875V and then change to 0.950V when the "profile/mode" changes to 450MHz Core / 1400MHz Mem / 0.95 VDD.

       

      I searched on the web but i didn't find nothing, except a way to check the BIOS values of the card, which are at the "correct" 350MHz / 150MHz / 0.875V for that mode.

      The Profiles/Modes in my BIOS (that i saved using the Trixx util that came with the card) are these:

      Core     /    Mem   /   VDD

      350MHz / 150MHz / 0.875V

      450MHz / 1400MHz / 0.950V

      1020MHZ / 1400MHz / 1.188V

      1070MHz / 1400MHz / 1.238V

      Except for the voltage in the first mode that is 0.950V, all other values are as they should be (as i read them using GPU-Z).

       

      I uploaded screenshots of anything i thought you would like to know here.

       

      I never changed/updated the BIOS of the card, nor ever messed up with overclocking it, although i did change the fan settings to Manual and then to 100% (all the time) since the summer kicked in, and that is a thing i am doing on at least my 3-4 latest cards.

      Winter temps were ~30c Idle and i think 66c is the higher i seen (I use HWMonitor to monitor the max temp my CPU and GPU reach when gaming).

      Now with 270X fans at 100% all the time, idle ranges from 36c to 47c while just browsing sites (core at 350-450MHz and VDD at 0.95V) and watching Youtube (Core at 1020MHz and VDD at 1.188V when in full screen 1080p). Gaming is much worst, i have seen 78c after a few hours of Dirt Rally, so i try to limit my gaming hours...

      Note that my Case is a standard Midi Tower with two 120mm fans in front and one 120mm fan back (plus the 120mm fan of the PSU that also blows out), but the fact that the card dumps the hot air inside the case (instead of out of it like my former Nvidia GTX275) does not help. CPU only reaches ~53c gaming now in the summer, cooled via a Thermaltake Big Typhoon cooler that also draws/dumps air inside the case like the 270X does...

       

      Driver version is 16.20 (More info in the screenshots).

       

      System:

      Asus P5Q3 Motherboard

      X5460 CPU (Xeon pin mod, think of a Q9650 instead)

      Crucial Ballistix 8GB 1600MHz RAM

      Corsair TX750W PSU

       

      Sorry for the long post but i wanted to give as much info as possible, and thanks in advance!

        • Re: R9 270X: 0.950V instead of 0.875V in 2D Mode
          black_zion

          78°C on a graphics card, especially after "hours" of gaming, is nothing to worry or even be concerned about.

           

          Are you using more than one display?

            • Re: R9 270X: 0.950V instead of 0.875V in 2D Mode
              alandolby

              Well it's not an alarming amount of heat sure, but it's not the best for the longevity of the card either. And i know that AMD cards tend to run a bit hot, but that's not the issue here really.

              I just noticed the 0.95V thing and i don't like things that are not "in order", not if i can do something about them at least
              And use one LG Flatron IPS235 23" monitor at 1080p.

                • Re: R9 270X: 0.950V instead of 0.875V in 2D Mode
                  black_zion

                  78°C is well within the normal operating range of every component of the card, it's not going to shorten its life appreciably.

                    • Re: R9 270X: 0.950V instead of 0.875V in 2D Mode
                      alandolby

                      "...78°C is well within the normal operating range of every component of the card..."

                      You are correct and i am assuming that 150c is the 'absolute maximum rating' value for most of the semiconductors on the PCB.

                      Having said that, i know from my background in electronics that higher heat is (almost) always a bad thing, and it has a negative affect on components, less on some and more on other:

                      The GPU chip itself will be fine with 25c to 100c+ cycles, the solder that joins it to the PCB not so much...

                      The Voltage regulating MOSFETS have no problem at all working at 100c for 10 or 100 or even 1000 hours strait, the electrolytic capacitors that are next to them and absorbing heat, will degrade faster because of that extra heat (yes even the solid electrolytes ones).

                      Degrading resin (because of high heat cycles) on high-power coils and transformers might not cause them to fail, but they may produce an audible noise because of that...

                       

                      In electronic design it's usually desired to have the lower possible heat (amongst other design factors ofc), even 60c rather than 70c in some cases is better (80c rated capacitors are cheaper than 100c rated ones). But its a compromise between things like cost, part count and such, so you are balancing the end design to meet all the specifications but be as cheap to produce as possible in most cases (i know of at least).

                       

                      I'm not trying to say that you should panic if your card is hitting 80c, i 'm not saying that a card that reaches 80c+ will fail before the warranty ends , although the latter is a calculating factor in a design and some small percentage of failing products makes 'financial sense' in some of the cases i have personal experience with.

                       

                      The first paragraphs of this reply and that part of the original post is just to explain my personal preference of having my electronics works as cool as i can, within again, my idea of what is a 'reasonable cost' (for example i would not use water-cooled heatsinks for my non-overheating PC).

                      But thank you for replying to my post sir

                       

                      So besides my personal preference for working temperatures, anyone knows why the card uses 0.950V when the BIOS value is 0.875V ?