There is no real fan stop, because what you have implemented here is very tricky. Instead of a real switch-on area, including hysteresis and start pulse, KFA2 simply relies on the inertia of the fans. I didn't want to believe it at first and measured it, but the card starts already in the fan curve with 2% PWM, where no fan is going away yet. This only happens at approx. 4% with approx. 170 to 185 rpm and then rises as normal. When cooling down, it goes exactly the other way, where it is simply controlled down and the fans simply stop by themselves at some point.
You can see very nicely on the curves that you set slightly higher speeds for the fans above the GPU and the in the middle (fan #1) than for the right outside (fan #2). In the stress test, a very similar picture emerges:
Let's look at the measured values again in direct comparison to the Founders Edition as a tabular listing:
| KFA2 GeForce RTX 2080 Ti Hall of Fame (Fan #1 / #2) |
GeForce RTX 2080 Ti Founders Edition |
|
|---|---|---|
| Fan speeds Open Benchtable Maximum | 1177 / 1070 rpm (Gaming, Peak) | 2136 rpm (Gaming, Peak) |
| Fan speeds Open Benchtable Average | 1150 / 1040 rpm (warmed up) | 2122 rpm (warmed up) |
| Fan Speeds Closed Case Maximum | 1296 / 1203 rpm (Gaming, Peak) | 2281 rpm (Gaming, Peak) |
| Fan Speeds Closed Case Average | 1274 / 1174 rpm (warmed up) | 2274 (warmed up) |
| Noise Emission (Air) Average | 38.1 dB(A), Closed Case (Peak) | 41.9 dB(A), Closed Case |
| Noise Emission (Air) Idle | Fan stop | 31.8 dB(A) |
| Sound characteristic / hearing impression | noisy, restrained, light engine noise | rather noisy, hardly lower-frequency shares |
| Coil-feathers/electric noises | Audible at lower fan speeds. | low, only for very high FPS numbers and for load changes |
Sound Spectrum
The measured 38.1 dB(A) are based on the measured 1150 (fan group 1) or 1040 (fan 2) rpm in closed housing. We have applied the same gaming load to the open body in the measuring room, but we have fixed the fans to exactly these speeds in order to be able to adjust the value as accurately as possible. The result is impressive, because the approx. 290 watts of waste heat want to be disposed of quietly and professionally.
Even overclocked, the whole thing remains acceptable and is hardly above the values of the Founders Edition, if you let the fans be controlled automatically. For approx. 370 to 380 watts of waste heat, i.e. 100 watts more, are only approx. 4 dB(A) more a really solid performance. In both settings you can see that KFA2 also acts a little with reason and consideration here, while some competitors howl as if there was no tomorrow. Fits!
Kommentieren