Page 9 of 9 FirstFirst ... 789
Results 81 to 84 of 84

Thread: HD 5750 IceQ has a bad design - Any news about Turks?

  1. #81
    Quote Originally Posted by DeathAndPain View Post
    To the contrary, you actually increase its life expectancy, because it stays cooler. This also reduces overall heat inside your computer case, making all of your hardware live longer. What rambaldi wrote is nonsense; there is no risk of damage from undervolting.
    In the range of voltages allowed on these cards under volting is completely safe but with out knowing the circuit designs used for power there could be conditions where power control circuitry doesn't act correctly, flooding the chip and burning something out. This is highly unlikely to happen in general with a properly designed chip where they took this into account. I don't know if they did though with GPUs as you wouldn't expect it to happen.

    The far more likely result of under volting is just to get glitches as you would from OCing with out raising the voltage enough.
    E-Peen: AMD Phenom II x4 965 (stock) | MSI 790FX-GD70 | ATI HD 5970 (stock) | Kingston HyperX DDR3 1333 CL7 2x2GB | Seagate 7200.12 4x1TB | Ikonic RaX10 Liquid

  2. #82
    Quote Originally Posted by Wandrey
    Although the GPU chip can endure high temperatures, one must take in consideration how the other board components are faring. Make sure the cooling configuration that let the GPU temp reach 90?C (which is indeed acceptable) is sufficient to cool the VRM and memory chips.
    Designing a graphics board for a GPU that is designed for 100+?C and then not layouting the rest of the board accordingly would be a poor performance on the side of the graphics card maker.

    Quote Originally Posted by Wandrey
    One more thing: GPU temperature is actually the temperature of the dispIO part of the die (at least on Radeons).
    In fact the Radeon chip has multiple sensors on its die. GPU-Z displays three temperatures, all from the Radeon chip. So they covered the whole chip nicely.

    Quote Originally Posted by Wandrey
    Finally, I dislike the auto-tune of CCC. I tested it only with a HD 4870 and it resulted in unstable clocks. I would like it better if it was a little more conservative.
    That is well possible (I did not experiment enough with it to tell for myself), and I did not advocate to go for auto-tune settings. All I said is that if auto-tune found 1240MHz to be working for my memory, I should not have any problems with 1150MHz.

    Quote Originally Posted by Nerdmaster
    Undervolting (from what I have heard so far) is considered safe when done properly but that is what rambaldi suggested after all.
    Please tell me how I can undervolt my chip in a manner that would not be safe...
    Problems worthy of attack prove their worth by hitting back! (Piet Hein)

    Core i5-4670, 16 GB memory (2x8GB Kingston HyperX RAM), Asrock B85M, AMD Radeon HD 7750, Creative Labs XiFi Titanium PCIe, Seasonic M12-II 430 Watts PSU, Samsung SSD 840, Seagate ST350041 hard disk

  3. #83
    8-bit overflow
    Join Date
    Jan 2010
    Posts
    582
    Quote Originally Posted by DeathAndPain View Post
    Designing a graphics board for a GPU that is designed for 100+?C and then not layouting the rest of the board accordingly would be a poor performance on the side of the graphics card maker.
    I believe that the board is designed to guaranteed endure the heat under default fan configuration. Slowing down the fan speed for better acoustic may or may not impact other components, even if it is fine for the GPU. The HD4870 is a good example of that, the VRM temperatures are the ones that should be controlled.

    Although I agree with you, I still think it is better to be safe than sorry, that's it.

    In fact the Radeon chip has multiple sensors on its die. GPU-Z displays three temperatures, all from the Radeon chip. So they covered the whole chip nicely.
    I said exactly that. There are 3 on-die sensors: one on display IO, one on shader core and one on memory IO. The "generic" GPU temperature showed by most softwares is the display IO one rounded down. The other sensors usually report higher temperatures, with the memIO one being the hottest.

  4. #84
    Senior Member
    Join Date
    Dec 2010
    Location
    Switzerland Geneva Cern
    Posts
    2,929
    Please tell me how I can undervolt my chip in a manner that would not be safe...
    Messing with the board to manually adjust voltages is not safe for example.

    But it would be dump to do that if you want to underclock since you dont need to drop the voltage so much.

    So yes for a cpu / gpu undervolting is safe.
    Current rig: i5 3570 @3.8GHz, Asus 7750, Dell 24'' 2560X1440, Corsair Vengence 16GB Ram DDR3 @1600MHz, Chieftec 550W PSU, 480GB OCZ SSD + 500GB Seagate + 1TB WD hd, Windows 10 64-bit

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  
WordPress Appliance - Powered by TurnKey Linux