Will we ever get back to a time where consumer CPUs can be passively cooled? Like they used to be in the Pentium 1/2 days?
underclock em
>>55487619
They can now.
If you don't mind throttling.
I don't see the point though.
CPU fans are quieter than case fans.
MacBook
>>55487619
trick is to get a big enough heatsink, such as
https://www.amazon.com/Prolimatech-Black-Megahalems-CPU-Cooler/dp/B005869KES?ie=UTF8&tag=pcpapi-20
>>55487640
Case fans are a meme. Hot air has a lower density than cold air which makes it leave the case on its own even without fans.
>>55487619
wait for Zen(tm).
>>55487952
>Hot air has a lower density than cold air which makes it leave the case on its own even without fans.
No heatsink here.
there are huge passive cpu coolers available
though even then it'd help a lot to at least have very slow fans moving air through fins/the case
>>55487640
>CPU fans are quieter than case fans.
wat
Beep boop
>>55488042
Not him, but hot air rises & dissipates quicker.
You're just dumb. Next time fart in your pants & time how long it takes for the stink to fuck off.
Dopey cunt.
>>55487619
i have a nh-c14 on a 6500k
When im watching movies i just plug out the fan header and let it cool passively.
never gets above 45c
total silence
>>55487619
are there coolers with like really slow fans that could cool down a 95W apu?
>>55488431
you know you can control fans in software on pretty much any remotely modern hardware, right?
just configure such software so that;
- it targets a highish maximum temperature
- the fans have a very low or zero minimum speed
>>55488057
it has small minifan
>>55488521
i tried but i have 2 3pin fans that i cant control trough voltage.. gigabyte mobo..
>>55488431
I have the nhl12and just set the fan curve to only ramp up over 55. Quiet at most of the time and much less hassle than unplugging can headers.
>>55488546
which model?
>>55488630
Gigabyte GA-Z170N-WIFI
>>55487952
> 'what is convection?'
>>55488696
hi, berendsp
>>55488696
Just wondering... What case and how high of an overclock? What temps under load?
>>55488732
fine detective work there
>>55488742
ncase m1
4.3gHz oc
temps under load with fans on max 60c
screenschot with prime95 running for 5 min
>>55488820
>fine detective work there
heh, well i looked around and didn't find much concrete information
you could try;
- updating the BIOS (http://www.gigabyte.com/products/product-page.aspx?pid=5518#bios)
- look around in the BIOS for fan speed controls, if the options it has aren't good enough, disabling them or setting them to "full speed" or similar may be needed to allow for software control
>>55487629
I really wonder, how much can you underclock a CPU?
Also, say CPUs like a 6300T — is it just an underclocked 6300? Do I get the same low power usage by buying a (considerably cheaper) 6300 and underclocking it?
I probably need a specific mainboard to do that, what do I need to look out for?
>>55488872
from what i understand, undervolting is what makes the difference
and underclocking is what enables further undervolting
>>55488888
check'ed
>>55488872
I don't know where the limit is, but my i7-4558U (cucked ultrabook dualcore) goes to like 1700MHz at 850mV in powersaving mode, I'd imagine you can go to like 1200MHz at 700mV with more recent 14nm CPUs
>>55488870
ye tried the bios, but no luck.
But i am using a low noise adapter so i don't even hear the fans when i'm sitting behind the pc.
its only when i watch a movie on the couch next to the cas that i can hear it in silent parts of the movie.
also checked the changelog of the new bios but no fan update. and i don't want to flash a new bios if i don't have to.
the ncase top panels pops right of i can just unplug the fan header.
>>55488920
If they're only 3pin then you can't control the speeds. The controllable chassis fans usually have a fancy "PWM" name as well ("be quiet! Pure Wings 2 PWN" are 4pin and controllable while "be quiet! Pure Wings 2" are 3pin and run at full speed at all times)
>>55488920
>checked the changelog of the new bios but no fan update.
in my experience they rarely list everything they changed
>>55488939
>>55488920
To add: Motherboards or even PSUs can't just adjust the voltage, it stays at 12V or 5V depending on the connector.
>>55488920
what case is that
>>55488955
disregard
>>55488939
>>55488952
well fuck don't feel like trowing another 30,- towards pwm fans.
>>55488941
only said updated cpu microcode
>>55488955
ncase m1 its a bit pricey but it looks really nice.
>>55488920
worst case you can always just get a hardware fan controller, which functions independent of the motherboard or software
http://www.ebay.co.uk/itm/DC-12V-Fan-Temperature-Control-Speed-Controller-CPU-High-Temp-Alarm-PWM-PC-CPU-/221878592436?hash=item33a8fea3b4:g:JwAAAOSwgNRV7qv1
>>55487619
Lots of atom/celeron mini-ITX boards ate passively cooled. I use one in a NAS. 38 degrees celsius under full load.
>>55487619
Get a big enough heatsink and a Pentium or i3, and there ya go.
>>55488380
>hot air rises & dissipates quicker
Yes but you still want to exhaust it, and having positive air pressure inside your case can help with dust management if you filter the intake spots.
For my sake, I put the 2 top fans outputting air, most of which is from my radiator set into the back (it has enough wall clearance), while the other 4 (one of which is 200mm) are all intaking it. The only time the fans are 'loud' is when I have the 2 on my radiator set at 2800rpm... which is all the time since I overclocked my CPU and started enjoying the hell out of Fallout 4 and some other gaymes.
>>55488978
what you can connect that to the pc and control the fans with software just like they are 4pin pwm fans?
and only 3 bucks..
might give this a shot
>>55487952
I studied marine cooling systems and thats bullshit.
The build up of hot air inside a case is faster than the dissipation of it, making the temperature rise as long as there is a source of heat. Having constant ventilation prevents the build up of heat.
Having no fans and a reasonable and stable temperature is only possible if the heat is lower than the dissipation.
>>55487643
this
just unplug the fan, throttling is no problem if you prefer silence to speed
What is Lenovo Yoga 900s?
>>55489000
What's bullshit? Studying marine cooling systems?
>>55489000
>making the temperature rise as long as there is a source of heat
Until it reaches equilibrium dumbfuck.
>>55488998
not with software
this device has a physical temperature probe which you would attach to your heatsink (near the bottom, where it's hottest)
and the DIP switches on it set what temperature you want to target (see the ebay description)
see the thing on the left? that's a temperature sensor (the other thing is a speaker, which acts as a warning if one of the fans fail)
>>55488998
You can't connect it to the PC in any meaningful way, you can cut off 12V rails off of molex from PSU to provide power for it, and just by looking at it, it'll only help you unplugging the fans easier (just by loosening the screw and letting a wire fall out (or you can solder a switch onto one of the wires from molex)), you still won't be able to control the voltage
>>55489019
case fans being a meme is bullshit
>>55489046
Indeed, I have 4 case fans
>>55489068
>4 case fans
You're reaching meme status though.
>>55487619
We already are, they're called mobile SOCs.
>>55489015
Haswell U capabilities. Guess that's okay.
>>55489119
The screen is absolutely atrocious though.
>>55488998
>>55489039
also, if you're wondering if 3 pin fans will work
this controller is only wired up to the first 2 pins (ground and 12v)
most fans work just fine when controlled by PWM, even if they don't claim to support it
>>55489142
3 pin controllers are better than 2 pin controllers though
>>55487619
???
>>55489183
like i said, worst case
this will work with just about any fan, and doesn't rely on support from the motherboard or software
still better than no fan control or manual fan control
>>55489045
>you still won't be able to control the voltage
this device controls the voltage going to the fans
>>55489245
No, it only has 2 pins connected
>>55489079
5 case fans here.
>>55489269
that's all that's strictly needed
>>55489275
Nigga, you just went full meme.
>>55489245
Huh... I thought it would just turn them on/off repeatedly to achieve whatever temperature... I didn't notice anything on the board that would change voltage, maybe I just don't know all my parts well
>>55489316
well yes, technically it turns them on and off
but that's what PWM is, it stands for pulse width modulation
the idea is that you pulse 12v at various widths (that is, varied ratio of periods where it's 0v and 12v) to achieve arbitrary lower /average/ voltages
>>55487619
Almost all pentium 1 and 2 had a fan
You want to go back to the 486 days
>>55489371
>>55489316
(and yes, the output from PWM is pretty noisy, typically the output is smoothed by a capacitor, but even then fans don't need particularly smooth power)
>>55489275
i don't think radiator fans should count
>>55487619
your phone contains one
>>55488140
how good is that?
>>55489275
>more than 1 outtake fan
do you realize you can't go below 0 pressure?
Are side fans a meme? I have an intake over my GPU.
>>55488939
Wut.
I have 3 pin fans connected to the motherboard and it adjusts fan speed just fine.
>>55488872
'T' typically isn't worth it, unless you have money to throw around.
They show wattage usage and it really isn't any usage.
New Skylake has real low power draw anyways. And Kaby Lake is supposed to be even better.
>>55491074
Side fans should cool the motherboard and gpu even better because its closer to them and blows air straight on them. There was some test and a side intake actually had slightly lower temps than a front fan system, everything else was identical.
>>55488920
>that power supply mounting
>>55491092
Depends on the board and the port. The 3-pin fan headers on my Biostar X79 board are locked to 12v, but the 3-pin fan headers on my older Asus P5Q-E, with the exception of the PSU fan header, all vary themselves from a common control signal.
>>55487619
Check out those passive netbooks on Intel Atom.
>>55489037
>Until it reaches equilibrium dumbfuck.
You mean in 9999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999 trillion years? Thank you for posting such practical information, I'm sure now we will just wait instead of buying a $4 fan.
>>55487640
The fuck do you mean, the loudest fan in my pc is my cpu fan
>>55487619
Why don't they just make CPUs huge instead of insisting on packing them with microscopic circuits that heat up like ovens?
https://youtu.be/etbh-YOQtp4
There's this.
>>55496169
because you can't afford more silicon
>>55496169
Because that means that electricity has to travel a longer distance inside the CPU, and electricity traveling through the silicon is what causes heat in the first place. Smaller CPUs put out less heat than bigger ones.
Rated for 95w TDP passively which covers the vast majority of Intel chips and around half of AMD's.
>>55496169
Less processors per wafer, which decreases yields.
>>55487619
You can do it right now, but you need to bring down your power usage a lot if it's a high end CPU and you need a special heatsink for it.
Like for example, high end heatsinks like Noctua NH-D15 won't be very good for it because they have really small fin spacing. I think a better idea would be to just under volt the processor a little and slap a huge cooler with a couple of slow speed fans on it. I had a Alpenföhn K2 with a single 140mm fan on it in my PC, running the fan at sub 800RPM at 5V. The PC is very quiet, specially with sound dampening. If anything makes a lot of noise, it's the damn hard drives.
>>55488742
Did that snek eat itself into a mini black hole or a rift even
>>55487619
> Pentium 2
> Passively cooled.
The last x86 desktop chip that could be passively cooled was the original Pentium 60.
>>55487619
10 cpu each running at 300-400 mhz each. Put it in one motherboard. Totally cool and silent.
>>55497864
my original comp had either a P1 133 or 166 mhz chip and it had only a heatsink on it.
I can passively cool with my NH D15 as long as I'm not really pushing it super hard.
Besides, the fans are totally silent like 80% of the time anyways, and if they aren't odds are my gpu is noisier anyways.
Would a fan blowing in from the harddrive area, one blowing from the roof and one blowing out from the back be an optimal setup for three fans?
If there's any extra spaces for fans without any installed in those spots, should I cover them up to keep a more consistent air flow?
>>55487619
>tfw new cpu cooler with massive heatsink and 2 140mm low rpm low noise fans
It's worth it brah. Silence is luxurious
Allow me to shill for a minute. The company I work for makes passively cooled computers. Some of the larger ones can handle 40w cpus. Pic related is our nicest looking system but it only dissipates about 15w.
>>55498996
But why?
>>55498406
It shouldn't have been. Normal deployment for anything above 80MHz required a fan. However, some OEM's did fuck huege heatsinks, because retarded.
>>55499014
Mainly for dusty and shitty environments. Try using a fanned computer in a woodshop.
>>55498996
That company doesn't make computers, it makes blunt weapons. I could probably smash a few hundred heads in with one of those things and it wouldn't even care.
>>55499160
We had a contest to see who could throw one the furthest at a company party. Shit was dangerous.
>>55498996
We have a few of your PCs at our warehouse. The coil whine is fucking ridiculous. I can play songs by plugging USB things in and pulling them out.
>>55499365
I'll bet they have the older 19v bricks we phased out. Testing 16 of those at once would give me a headache. All the new 12v stuff is silent.
>>55487619
that's core m doe
>>55487643
OP never said passively cooling CPUs at the expense of performance.
