Performance Test

The following software has been used to carry out the tests::

• UnigineValley Benchmark 1.0
• Battlefield 4

For each test we have recorded the data with Aquasuite before, after and during each test; then we have constructed graphs; we have taken room temperature and the temperature inside our Enthoo Primo case into account and adjusted the maximum temperatures that were recorded.

We run a Unigine Valley test for 30 minutes and in the following graphs you can see the highest temperatures recorded both with the default clock frequency and with the overclocked profile, with a GPU core clock at 1287 MHz (with a boost clock during the test at 1438MHz); this has brought an increase in voltage of about 87mV and of 125% for the power limit; this test has been conducted to stress especially the video card VRMs.

It is useful to remember that for the second overclock test, we have set manually the fan of our reference GTX 980 VGA at 100%.

As can be seen from the tests, Aquacomputer Kryographics GTX 980 yields excellent performances! In particular, with the overclock profile, the GPU core temperature does not increase, still leaving wide margins. As a matter of fact, we pushed even further the card and we managed to close several benchmarks and even play for a few hours under a 1551Hz Kepler-Boost profile without exceeding 50 °C!
Obviously, we cannot make comparisons with air cooling systems, especially stock coolers. However, we can point out that there is a difference of 25°Ceven with the manual setting of the stock heat sink fan at 100% (with obvious consequences on the noise),whereas obviously Kryographics GTX 980 operates in complete silence.

Finally, we have tested the card in a 90 minute session with Battlefield 4, aiming at simulating a daily scenario as close as possible to that of the average user, who clearly does not spend his days performing heavy stress tests or benchmarks. For this reason, we have also decreased the fan rotation speed to 1000 RPM and the pump speed down to 75%, in order to reduce noise, although experiencing a small fall in the performance of the cooling system.

In this test too, as in the previous ones, Aquacomputer fullcover performs greatly even with a lower flowrate. Obviously, the comparison with air stock can only give an idea of the general power of liquid cooling. However, in the near future, as we have already done for Nvidia GTX 780 and AMD Radeon R9 290, we will test other fullcovers, in order to make a comparative analysis of liquid cooling systems only for this GPU. Stay tuned!