Acoustic and Thermal Results
Whilst there was some coil whine in the menu screens and some high FPS benchmarks, it was probably the best card we have tested in recent memory with regard to the coil whine phenomenon.
The Twin Frozr VI cooler is so quiet that we struggled to get a consistent reading during testing over ambient room noise. During the Futuremark TimeSpy Demo, we managed to record a stable 21dB(A) at a distance of 30cm from the open air test bench. The previous Twin Frozr V cooler was already a champ in our minds and we were wondering if the MSI design team would be able to top it but the results speak for themselves. It’s quiet, effective and has more lighting bling – so, yes, it is better than the last one.
When we moved the test bench into a Fractal Design Define R5 case, the MSI GTX 1080 Gaming X was silent unless playing an intense game or running benchmarks, at which point it was barely audible from less than a metre away.
The card itself stays under 71C but doesn’t vent much air out the back of the card due to the 3xDisplayPort, 1 DVI and 1 HDMI video ports. This means that the majority of the exhaust ends up inside the case. With slow but steady case airflow, the case ambient temperature was easy to manage which meant that the CPU and GPU coolers didn’t have to work too hard.
Power consumption for the whole X99S test system was ~107W at idle and ~304W when running Unigine Heaven.
We transferred the X99S test system to the Fractal Design Define R5 chassis, keeping all components except the Optical Drive (to be honest, I couldn’t be bothered moving it and we never use it anyway). The MSI GTX 1080 Gaming X went in the case without any dramas, hardly made a peep while running and gave us the quietest high-end system we’ve used. The only regret was that the Define R5 we have here has a solid side panel (i.e. no window) so we couldn’t see the card inside. Everything lined up perfectly with the rear PCIE slots and we didn’t notice any flex in the card despite it weighing over 1Kg.
GPU Boost 3 has changed the way the new GTX 1080 and 1070 cards are overclocked so we needed to grab the latest version of MSI Afterburner to overclock the GTX 1080. As this is a pre-overclocked card, our expectations were not overly optimistic. With the fan running at almost full speed, we were able to hold a steady core speed of 2076MHz which gave us an average performance increase of about 8-10% in the synthetic benchmarks. Interestingly, this ‘extra’ overclock made no real difference under the in-game benchmarks.