Real achievable clock rates and overclocking
I can summarize that with the clock rates and overclocking. The GPU clock of the overclocked card briefly goes up to a whopping 3075 MHz when cold, then levels off a few boost steps lower at around 3030 to 3045 MHz more or less alternately once fully warmed up, which is still a lot. That can definitely be left alone. With manual OC of the memory you also get the min FPS a bit nicer, but that was not the purpose of our OC experiment.
GPU temperatures in the case
The differences are not as big as feared if you use a proper case. If you close the panel, the GPU temperature (Edge) increases by 3 Kelvin, the GDC hotspot increases by up to 4 Kelvin. Nevertheless, the maximum values of 62 °C (Edge) and 74 °C (Hotspot) are nothing to be afraid of. The smaller delta compared to the Palit card results from the real heatsink and the three heatpipes compared to the cheap DHT cooler (Direct Heat Touch) from Palit, where a part of the GPU only rests on the aluminum heatsink.
Infrared measurements – Thermal Imaging
We still remember the uncooled voltage converters on page 3 in the teardown, where the airflow has to fix it. And yes, that actually works quite well. In gaming, however, we see that a second MOSFET in the memory could have lowered the temperature a bit, but that also costs money…
In the stress test, the upper voltage converter for the GPU is the hottest. And we can also see the individual layers of the board very nicely, which is otherwise not so clearly visible. All in all, it’s a good result and if you want, you can slap two ultra-soft 3 mm heat conduction pads on the two neuralgic spots.
Fan curve and operating noise (“volume”)
Now we come to what automatically arises when the waste heat is removed: the good old fan noise. However, the card pushes the fan hard at first, only to level off again a bit further down. That’s quite good at first. In the load scenario, this is then around 1200 rpm, which is barely audible. However, we see a temporal offset between the temperature and the fan curve, which indicates that the switch-on point was set a bit too high.
But what does the measurement for the MSI GeForce RTX 4060 Gaming X 8G reveal then? With 31.2 dB(A) in the gaming loop and plenty of 135 watts, none of this is an issue, although we can hear (and also see as slight peaks) the minimal harmonics of coil and capacitor buzzing. But it’s not really much this time.
Coil Noise: “Coil Whining”
Therefore, we still have to talk about one point and that is once again the annoying coil whining, even though it turns out pleasantly low here and is only audibly noticeable at FPS rates above 400 FPS. I already wrote that the card spits a bit in the load peak measurements. Especially since the five phases have to work hard. More about this topic can of course be found here:
- 1 - Einführung, technische Daten und Technologie
- 2 - Test System im igor'sLAB MIFCOM-PC
- 3 - Teardown: PCB und Komponenten
- 4 - Gaming Performance FHD (1920 x 1080)
- 5 - Gaming-Performance WQHD (2560 x 1440)
- 6 - Gaming Performance DLSS vs. DLSS3 vs. FSR
- 7 - Lastspitzen, Kappung und Netzteilempfehlung
- 8 - Temperaturen, Taktraten, Lüfter und Geräuschentwicklung
- 9 - Zusammenfassung und Fazit
20 Antworten
Kommentar
Lade neue Kommentare
Urgestein
1
Urgestein
Veteran
Mitglied
Urgestein
Veteran
Veteran
Veteran
Urgestein
Urgestein
Veteran
Mitglied
1
Mitglied
Moderator
Neuling
Neuling
Urgestein
Alle Kommentare lesen unter igor´sLAB Community →