Ask HN: Why aren't laptops evaluated with a plot of performance vs. fan noise?

55 points by amichail ↗ HN
Nobody wants to use a laptop that has high performance but also fan noise for extended periods of time.

43 comments

[ 2.8 ms ] story [ 81.6 ms ] thread
Gamers Nexus has done such testing for GPUs, dB adjusted performance numbers. No reason you can't for laptops.
I like the reviews at https://www.notebookcheck.net/ because they do measure fan noise.
The fan noise and laptop hot zones are two of the reasons I always check that site. Many laptops will serve as an adequate substitute for long term birth control if you actually use it on your lap.
Or a replacement for room heater. Proud owner of HP Pavilion DV5-1106AX which got me through winters of 2008-12 when I was broke.
(comment deleted)
NotebookCheck[1] does extensive fan noise, chassis heat, and hot air distribution tests.

It also performs display colorimetry and calibration, display PWM tests, Wi-Fi and SD card slot (if present) speed tests, basic teardowns, extended performance tests (accounting for PL1/PL2 limits), combined CPU + GPU performance tests on a veritable battery of benchmarks from games and renderers to industrial benchmark suites on a variety of resolutions and settings.

[1]: https://www.notebookcheck.net/

Thank you and sibling very much for that recommendation! They l show noise levels per frequency band.

I skimmed several reviews on that site in the past but never saw that. Reason is they were their equally great mobile phone reviews, which i read to check for blueish tint of displays.

For GPU noise comparison, techpowerup.com does it best. Pick the latest test of the model you are interested in -> Go to the Noise emissions page -> Directly compare against all other boards with that chip they measured.

They are great, but they don't review many laptops.
Apple MacBooks: what fan noise?
Macbook Air performance / fan noise...

Error

Was sitting between two Intel macbooks the other day, like sitting between two hoovers, they haven't all gone away yet!
The last Intel MacBook was discontinued 16 months ago
There is not specific reason I think.

But consider that a lot of people live in environments so loud that for a lot of laptops the max fan noise is at best a mild nuisance, at least with modern laptops.

Also a lot of laptop gaming is done with headphones.

Then there is the problem that measuring fan noise is harder to do then measuring performance, for the first you just have to run benchmarks in a "normal" temperate room, for the later you need a consistent not too loud noise environment.

IMO title is clickbaity; it's a statement-of-opinion-masquerading-as-a-question.

Effectively the parent is saying 'fan noise is up there with performance as my most important metric in a laptop'.

This may be true for the OP, but it's by no means a universal consideration. Personally, battery life comes before fan noise for me as a metric. YMMV.

To a first approximation, if the fan is going (and using energy) because the CPU is running hot under load (using energy), doesn’t that give you a hint about battery efficiency?
Two laptop model with the same CPU can have big differences in thermal design, which impacts a lot the fan noise but battery efficiency only in a marginal way.
Sounds like you’ve just supported the point about fan noise being a good approximation for battery efficiency.
"Nobody"?

How much "fan noise"? Compared to what? An anechoic sensory deprivation chamber? A jack hammer outside the office building? Your refrigerator? Car tires at highway speed?

> How much "fan noise"? Compared to what?

It doesn't need to be "compared" to anything. Just a chart of noise in dB vs. performance would suffice.

That way you can glance at the chart and say "at this performance level, I can expect it to make this much noise"

It does need to be compared to something - 72dB is "loud" in a quiet room

It's pretty "quiet" in other contexts

It can just be measured and reported, and the user can decide for themselves?
Honestly, almost no one understands decibels ... beyond "less if better" (for some things) :)

I took a whole semester class in college on the physics of sound - decibels and loudness are just bleeding weird :)

You don’t really need to compare it to anything other than the 0dBSPL reference point.

Standardize a laptop setup and environment, pick a listening position, measure dbSPL. A good position might be “1m from the center of the laptop base, at an elevation of 45°, while the laptop is on a flat, hard desk”. You know, more or less how people use laptops. Obviously you can go loose or specific with the description of the listening setup. You could standardize the room, you could standardize the desk, etc. I think the desk probably should be there, since it will reflect fan noise, and closer match actual laptop usage.

You measure the A-weighted fan noise under different loads and publish something like dB SPL versus W graph.

> How much "speed"? Compared to what? A human-powered recumbent bicycle? A diving peregrine falcon? Your lawnmower? A train?

Much like speed, we have standardized units and measurement methods for sound.

>How much speed?

See my example: "highway speed"

I do not care about fan noise because I am always listening to music. I'd rather have much higher performance for louder fans.
?

I do care about fan noise because I am always listening to music.

(comment deleted)
Similar sentiment: who cares about laptop fan noise when I'm already wearing noise cancelling headphones?
There are other variables

- Room temperature: It makes a difference whether it is summer or winter.

- Throttling ratio : A lot of laptops will lower the CPU speed rather than keep the fan running for a prolonged time, or am I wrong ?

silentpcreview.com abandoned their mission :-(
Because most people gaming on laptops are using beefy headphones.

Quiet operation is a luxury anyway. You don't buy a laptop to be quiet, you buy a laptop to perform the workloads you need (for which you need performance). People who are budget constrained usually want the best possible performance for the price.

I buy a laptop to be quiet.

If I can't get a quiet one, I don't buy that laptop.

If no laptop performs the workload I want without noise, then I use a quiet desktop, or remotely connect to the loud machine in another location.

Computer review methodology is terribly anti-scientific in general. For example, look at all the pomp and circumstance around reviews of desktop PC coolers - we used this motherboard and this RAM with this case in a room this temperature blah blah. Whereas the physical reality is that thermal resistance (degrees temperature rise per watt dissipated) sidesteps all of that. The right way to characterize a cooler would be to use a constant heat source to measure the thermal resistance (via temperature) and noise while varying the fan RPM, and create a simple graph that could be straightforwardly compared. But then they wouldn't be able to focus on a subjective narrative supporting whatever manufacturer paid them to write the article.
hey. anyone wanna help me?

i bought a used laptop "lenovo ideapad" ryzen 3 something.

Am running kde neon. The only problem, the FANS ARE 100% ALL THE TIME.

it gets tiring pretty fast because the jet blower of the fan is supposed to come up during heavy load but not all the time.

i tried pwmconfig but it couldn't find fans. Same for other utilities. They just don't recognize the fans.

Could it be a hardware issue or a software one beacause the fans are working fine, they just don't modulate.

I'd much rather see sustained performance under peak load with respect to power throttling. A laptop with a 4 GHz CPU is worthless if you can only run at that clock speed for bursts of a few seconds at a time.
A powerful CPU is not worthless if you only run demanding things every now and then.

I bought a powerful CPU not because I wanted to calculate pi continuously all day long, but so when I did do something demanding it was able to do that in a reasonable time.

99% of the time my CPU is running <10% usage. Same with most people I'd think.