Basically just the CPU fan, case, and PSU (if you’re lucky). If someone can figure out how to get Doom running on just those 3 things, young people will be able to ascend.
Basically just the CPU fan, case, and PSU (if you’re lucky). If someone can figure out how to get Doom running on just those 3 things, young people will be able to ascend.
You were downvoted because you’re wrong, or at best technically accurate but inaccurate about the results.
Your fan speed is set by your fan controller.
And while it’s true that different fans can be set up to spin at different speeds with the same input voltage or PWM signal, if the fan is less effective at cooling at that rate, the controller will end up higher on the fan curve and speed it up to compensate.
The only things that determine your fan’s ultimate speed are a) how high the controller allows it to go, b) how much heat the hardware is generating, and c) how efficiently the combination of coolers and fans dissipate that heat into the air (which also depends on the room’s conditions).
Now we’re just talking about fans, so we can ignore item b entirely as well as the cooler and room conditions. And it then comes down to, can the fan keep up with the heat being generated at the current point in the fan’s curve? If not, heat increases and the controller speeds up the fan. Then, when it hits whatever the cap is, if it still can’t keep up, the hardware either overheats or throttles.
If it can keep up, then it stays where it is on the curve. If it can dissipate more heat than is being produced, it goes down on the curve and the fan slows down as the temperature lowers.
All this to say, if you have fan A that needs to be at 2000 rpm to keep up with the heat being generated by the hardware at say 65°C and someone says they have a better fan, B, that is just set to spin slower at the same signal from the controller, all that will happen is it will fail to keep up with the heat generated at 65, then the heat will rise until it’s spinning at 2000 rpm anyways or the curve hits the cap and the system throttles to reduce the heat generated.
So we couldn’t really eliminate that item b above and it could actually be relevant, though if that were the case, I’d assume people would just curse noctura because their fans suck and their systems run hotter or slower. But they have a good reputation where the main complaint I see is about their price, not their performance.
Personally, I use be quiet! fans rather than noctura, but it’s a similar principle. They make quieter fans by tweaking or redesigning blades shape and angle, bearings, and the motor. Blades make them more effective at slower speeds, and all three can make them quieter at the same speeds.
I disagree about why I got downvoted, but I don’t care enough to argue the point.
You’re correct about the temperature/duty cycle curve, but not “the only things that determine maximum fan speed” and what that means in relation to that curve. I’m loath to toot my own horn, but the concepts you’re describing and are claiming I don’t understand are what I work in professionally. I mentioned earlier I’m an EE - an electrical engineer.
Maximum fan speed is inherent to the fan itself and, put very simplistically, in a brushless DC motor is determined by the torque generated. As the motor spins faster, the total resistance (again simplistically, back EMF and net friction) increases, and once it equals the motor’s torque output, the motor has reached its maximum speed. Fans operating at a given voltage will have different maximum speeds based on blade size and geometry and motor design.
PWM fan controllers simply correlate a temperature with a duty cycle percentage, but cannot cause a fan to spin faster than its maximum speed. If one fan is rated for 3500 RPM and another 6000 RPM, they will spin at their respective rating at 100% duty cycle. No more.
Noctua’s fans in general tend to have engineering superior to many lower priced alternatives, somewhat reducing the sound generated at a given speed, but the low-noise line of fans for which they are often noted achieves that low noise primarily through reduced speed, and thus reduced output. Similarly, the low noise adapter included with some fans is simply an in-line resistor, also reducing speed for reduced sound.
My main point though is that it isn’t the electronics or the fan’s max speed that are the critical parts, it’s whether the fan is able to dissipate the heat being generated with the total sum of how everything is set up. If it can’t, then temperature creeps up until that max speed is hit (thus it can’t dissipate anymore and throttles, hits equilibrium at a higher temperature (because bigger temperature differential means more heat transfer even at the same fan speed, so it’s possible it won’t just run away when the max speed is hit), or melts/overheats to failure) or until the fan speeds up and is able to find an equilibrium or reduce the temperature.
Thus whatever tricks they use need to result in fans that are still good enough to replace whatever fan they are replacing or they will cause performance issues.
Though if someone is just adding case fans, then as long as it doesn’t disrupt the airflow, it will help even if it isn’t very good, and you can use multiple quiet fans to replace a single loud one and see an overall improvement even if the individual fans aren’t nearly as good as the single one they replace. So based on that, I’ll back down from my previous “their reputation speaks for itself” because the reality is that they can be overpriced underperforming garbage but still get a good reputation because they aren’t just being evaluated via a/b testing and I was naive to oversimplify it like that.