HIS HD6970 2GB IceQ MIX (Lucid) Review

👤by Richard Weatherstone Comments 📅13-07-11
Overclocking & Temperature Analysis
Overclocking



As you can see above, the stock card weighs in at 880MHz on the core and 1375MHz on the memory. This is attained with a VCC of 1.175. Using the Afterburner tool from MSI I began by going for broke, raising the Core Clock and Memory clock options to their maximum values (950/1450). Unfortunately, this made the setup lock up almost instantly as soon as any 3D application was started.



Being a little tamer on the overclocks and spending more time with the overclock settings, I achieved a respectable extra 35MHz on the Core and 52Mhz on the memory with no voltage increase. This wasn't a bad result but I still felt the card had much more to offer so I raised the VCC from 1.175 to 1.299 which allowed the maximum overclocks of 950Mhz on the Core and 1450 (5800MHz effective) on the memory, an excellent result! Sadly this was as far as Afterburner or Overdrive would allow me to push the card but I have little doubt clocks could be tweaked even further with a little BIOS editing as esentially, the card is the same design as the Turbo Edition which breezed 1000/6000MHz.

Temperature

To measure the temperature of the card we first measure the ambient temperature. Our review station resides in an air conditioned room with the temperature set to a comfortable 23c.



In stock, idle state, the GPU temperature is a hovers at 44c. Not a great results given the size of the cooler but at idle, the GPU is silent. This is with the GPU in its power saving state running @ core:250mhz, memory:150Mhz with a power draw of 4W (0.900v).

Placing the GPU under 100% load and running at its factory settings of 880/1350MHz (Core/Memory) we see the temperature steadily rise along with the fan speed. The highest temperature after 5 minutes reached a plateau at 75c. This was with the fan running automatically (66%) at 2121 rpm.

With the card fully overclocked using the optimised settings found in the overclocking above, the temperatures were slightly higher as expected, settling at a maximum 85c after 5 minutes using the 'Burn-in test' of Furmark, returning to an idle 45c in a matter of seconds once the test was complete. In the interests of completeness, I let the Burn-in test run for a further hour and I am happy to report that the temp did not budge from the same 85/86c although the air conditioning in the room was working a little harder that before!

As Furmark was used for this series of tests it is unlikely that your GPU will exceed these temperatures under normal gaming conditions. As such, these results should be interpreted as a 'worst case' scenario and are testament that whatever your game, if the card can cope under the immense stresses that Furmark has to offer, it will be able to compete with any gaming scenario.


21 pages « < 9 10 11 12 > »

Comments