Fan speeds
In gaming mode, the MSI RTX 5090 SUPRIM LIQUID SOC shows a differentiated behavior between radiator fan and shroud fan. The GPU reaches its temperature plateau at around 53 °C, with the fan on the radiator remaining at around 1250 rpm after a short transient response. The integrated shroud fan, which primarily ensures local ventilation of the VRMs and the power supply, runs constantly at around 1300 to 1350 rpm. Despite the parallel activity of both fan groups, the background noise is quite perceptible, whereby the dominance of the air movement through the close-meshed radiator grille is more significant than the low-frequency noise component of the shroud fan.
In silent mode, the operating parameters shift significantly in favor of lower noise levels. The GPU now reaches around 56 °C, albeit with noticeably reduced fan speeds. The radiator fan drops to around 950 to 1000 rpm, while the shroud fan operates slightly below 1250 rpm. Both curves show a stable and smooth profile without any significant oscillations or abrupt readjustment, which indicates a conservatively clocked fan control with extended hysteresis parameters. The temperature increase compared to gaming mode remains moderate at around 3 °C, which underpins the efficiency of the semi-passive control.
The subjective noise characteristics are difficult to determine, however, as there are two overlapping sources with different sound radiation characteristics. On the one hand, the airflow from the radiator fan is clearly audible as a whistling sound at higher speeds, while the gentle whirring of the shroud fan dominates in the lower load range. It is therefore difficult to define an objective overall emission, especially for housings with variable acoustic damping and depending on where the radiator is installed.
Overall, the silent mode proves to be surprisingly efficient, provided there are no thermally limiting conditions such as summer operation or enclosure inclusions. The temperatures achieved are still in the uncritical range even at lower air flow rates, with the speed reduction enabling an acoustic gain without the temperature level drifting into thermal throttling. The choice between the modes can therefore be made according to acoustic preference rather than thermal necessity.
Noise emissions (“volume”)
The noise development of the MSI RTX 5090 SUPRIM LIQUID SOC is made up of two dominant sources: the fan on the radiator and the integrated pump in the cooling block including the fan in the shroud. Although both components are thermally efficient, they generate a differentiated acoustic profile that is difficult to capture objectively in a single measurement. The reason for this lies in the spatial separation of the sound sources and their different frequency characteristics.
While the radiator fan tends to cause broadband flow noise due to its larger air volume, the pump operates with a clearly defined tonal component that can be perceived as a humming noise, particularly in quieter environments, and significantly influences the measured value. In many cases, the pump is closer to the microphone or the user’s ear, which can subjectively reinforce the impression of a higher volume, although objectively less sound pressure is emitted.
In synthetic load scenarios, a combined sound level of around 34.5 dBA was measured in gaming mode, which is in the range of quiet case fans. The card thus remains acoustically inconspicuous in typical gaming situations, provided no open test environments are used. In silent mode, the noise level drops slightly to around 33.2 dBA under continuous load, which initially seems pleasant, but can be explained by the only slightly increased pump speed and a significantly less aggressive fan curve overall. The lower fan speed in the Silent profile ensures a softer modulated noise characteristic, but at the same time the pump is acoustically more prominent without being perceived as really annoying.
Such measured values should always be interpreted in the context of the measurement environment, microphone position and subjective perception. In a closed housing with decoupled radiator mounting, the difference between the two operating modes is significantly reduced. For users with sensitive noise perception or an open setup, it is advisable to define an alternative fan curve in order to better adapt the acoustic balance to personal preferences.
- 1 - Introduction, overview and technical data
- 2 - Test system and equipment
- 3 - Teardown: PC, components and cooler
- 4 - Material analysis and TIMs
- 5 - Gaming performance
- 6 - Power consumption, transients and PSU recommendation
- 7 - Temperatures und clock rates
- 8 - Thermal imaging and modding
- 9 - Fan curves and noise
- 10 - Summary and conclusion
30 Antworten
Kommentar
Lade neue Kommentare
Veteran
1
Veteran
Moderator
1
Urgestein
Urgestein
Urgestein
Urgestein
Urgestein
1
Urgestein
Veteran
Urgestein
1
Urgestein
Urgestein
Urgestein
Mitglied
Alle Kommentare lesen unter igor´sLAB Community →