At home in the air and underwater: Asus GTX 1080 Ti ROG Poseidon Platinum in test

All of this sounds simple at first, even in practice. Instead of a water block, simply place a flat heatpipe on a large heat sink and optionally let water pass through to absorb the waste heat and transfer it. Asus uses its own design for this board. The two 8-pin sockets for the external power supply are followed by two coils in the input area for smoothing the tips. Asus relies on a design of 5+2 phases, with the 5 phases for the GPU duplicated... Benchmarks in 2560 x 1440 pixels We have deliberately dispensed with Full HD (1920 x 1080p) as the card runs into the CPU limit even in the highest settings. In WQHD (2560 x 1440 pixels), on the other hand, the actual working environment and the kar... Benchmarks in 3840 x 2160 pixels The card is also significantly faster in this high resolution than a GeForce GTX 1080 Founders Edition or GeForce GTX 1080 Founders Edition. TitanX (Pascal) in Nvidia reference design. Overall, many titles are quite good in UH... Power consumption at different loads We measured the power consumption of the air-cooled map. When operated as a water-cooled card, the power consumption at full load is approx. 3-5 watts lower (fan). Compliance with the main... Overclocking The overclocking with air cooling is almost hopeless, because the card is already quite neatly overclocked from the factory and thus reaches its physical limits, especially in the closed housing. With a good water cool... Cooling system and backplate The special feature is the use of a hybrid system that can be operated both in a water cooling system and only alone with air cooling. The advantage is that the buyer is... Summary There is no egg-laying woolly milk sow, even with graphics cards. After this test, we must hold this honestly and fairly. The Asus GTX 1080 Ti ROG Poseidon Platinum has no serious flaws and does nothing ...

Overclocking

The overclocking with air cooling is almost hopeless, because the card is already quite neatly overclocked from the factory and thus reaches its physical limits, especially in the closed housing. With a good water cooling, even 2076 MHz are in it, but our chip was already at this point already at the end despite chiller, power target increase and small voltage injection. GPU lottery.

Temperatures and clock rates

First, we compare the achieved start and end values for temperatures and GPU clock (boost) in tabular form:

	Initial	Final value
Open Benchtable
GPU Temperatures Air Cooling	40 °C	76 °C
GPU Clock Open	2012 MHz	1924 MHz
Ambient temperature	22 °C	22 °C
Closed Case
GPU Temperatures Air Cooling	42 °C	78 °C
GPU Clock Closed Case	2000 MHz	1898 MHz
Air temperature in the housing	25°C	38°C
Water Cooling (Chiller)
GPU Temperatures @stock	22 °C	35 °C
GPU clock @stock	2025 MHz	2000 MHz
GPU temperatures @max. Oc	22°C	37°C
GPU clock @max. Oc	2076 MHz	2076 MHz
Water	20 °C	20 °C

Overview Graphs: Temperatures vs. Clock

For better illustrations now again the respective courses considering our timeline of a total of 15 minutes each for the warm-up time.

Thermal analysis of the back of the board

Finally, we look at the thermal analysis of the respective load characteristics. While the base temperatures and the GPU internal temperature are identical in air cooling due to the slower heat dissipation, the water cooling shows a small weakness of the cooling system used. While a normal fullcover water block allows no more than 5 Kelvin between water and GPU diode, the hybrid solution allows up to 17 Kelvin in the OC. In addition, the board temperature below the package is now also significantly higher than the GPU internal temperature.