if AMD and intel are so smart, why dont we have this?

[Confirmation]

if heat goes UP, and our tower cases are always vertical

why dont we have the PCI slots where the MEMs are, and the MEM ( lowest heat ) in the lower part with hdd, up of hdd the northbridge, up the northbride GPU, and cpu left side

WHY? 20 years in the market and they never changed this? -.-

 

[Fanatical Meat]

if heat goes UP, and our tower cases are always vertical

why dont we have the PCI slots where the MEMs are, and the MEM ( lowest heat ) in the lower part with hdd, up of hdd the northbridge, up the northbride GPU, and cpu left side

WHY? 20 years in the market and they never changed this? -.-

Requires smaller video cards maybe?
Also I believe this stuff is industry standard so changes take forever.

 

[Confirmation]

Requires smaller video cards maybe?
Also I believe this stuff is industry standard so changes take forever.

if all the heat goes up, our motherboards are wrong designed, even if you put the case in a server rack, im thinking that my position in that image, is close to how any case has to be, to dont make the gpu give the heat to CPU, and the psu give the heat to the gpu + cpu

basic abc is, lower power consumption components should be in the lower part of the case, and the highest components in the highest part " if heat goes always up "

also in SLI, always one GPU will give a lot of heat to the other one, if hot air always goes up, is like the first one will take the coolest air, and the second one " up " will take the hot air, and this is the normal design in SLI, i think it is wrong too, and watching my middle case,im thinking that with a good mother board, and case things, it could be fixed, i can buy some external pci cables, but, if only my MB were well designed, i wont have to do that

then in 20 years no one detected this? or whats the problem of try another design in the mother board than the basic

 

[Confirmation]

this is my trash solution to my case and temps, but is ugly, and i wont do this of course..

FIXED -->

NORMAL --->

believe or not, in a closed case, where you protect the components from the dust, changing the air flow in this way, is really great...

 

[NostaSeronx]

Look up Silverstone's Raven. Everyone is just building ATX cases wrong.

 

[Confirmation]

Look up Silverstone's Raven. Everyone is just building ATX cases wrong.

found him, but not found the video talking about the airflow, can you send me one? thanks

 

[Confirmation]

its like my cpu sends hot air, and my GPU gets it.. in the normal case

with a gigabyte could look more logic for my cpu heatsink

 

[Campy]

Heat doesn't go up, it goes where the fans make it go.

 

[Confirmation]

Heat doesn't go up, it goes where the fans make it go.

yes, thats true, but in a closed case, can make a hot air bubble, as is happening with my build, even putting fans in the corners, i found a way to fix it, but that way, demands change some things in the case, and i modded it in the past

in the first one build i have the issue playing gta IV that temps went to 90 degrees...

changing the air flow, doing some news air access, in a closed case, finally i got 20 degrees temps playing 1 hour or more

but that test where when i had a 9800 gt, when i got the new cards, air hot bubble started to happen again

i know sounds like a dumb thing maybe what im saying and i can fix everything with a water coolers

is just about i want fix it in air, and not pass to water

maybe i just had bad luck, a bad cpu silicon, and a bad gpu silicon, thats probabbly, as usually, to my country, comes the worst technology, is not the same buy an i7 920 in my country, and buy it in united states

in united states will be first hand

in my country, second or third hand, and market guys of my country will be buying it cheap from international store, and expensive in my country " is what happen " also you have stores in some countries, checking the silicon of the cpus, what cpu uses less voltage and less temps, and things as that, and selling it cheaper or expensive

 

[VirtualLarry]

Get cases with top vents / blowholes. I had one in earlier years, lowered my temps quite a bit.

 

[DrMrLordX]

1). dGPUs are optional. You'll find that the vast majority of PCs sold today don't even have dGPUs in standard ATX tower cases.

2). It's easier to do as @VirtualLarry indicated by changing airflow dynamics in the case than it is to try and draft an entirely new ATX spec for the tiny minority that still uses hot dGPUs in PCIe slots.

The rest of the world has more-or-less moved on from ATX spec PC tower cases. Take a look at notebook, NUC, and server room/blade cooling designs to see what I mean.

 

[NTMBK]

ATX is a rubbish standard, because it was designed around 90s CPUs and PCI cards that did not use much power. But it's too widespread to easily replace now. Intel tried to replace it with BTX, so that they could better cool their power hungry Pentium 4, but nobody wanted to buy it. They wanted ATX cases to go with all the ATX motherboards, and ATX motherboards to go with the ATX cases.

 

[DrMrLordX]

But it's too widespread to easily replace now.

I dunno about that. The real issue would be the cost involved, since the ATX 2.0 market is actually pretty limited now. Who wants to retool board/case designs at what would likely be a fixed upfront cost for a low-volume segment?

 

[IntelUser2000]

the market and they never changed this? -.-

It's not a technological limitation. The market has been using the same form factor for few decades now. You can't easily switch away from this.

You don't switch a widely established standard for benefits for a small subset of the population, because that'll inconvenience the rest.

Technology serves us, not the other way around.

 

[SirCanealot]

I believe Gamer's Nexus have flipped a few cases to see if heat=rises makes any difference, and I don't think they've really seen anything themselves. I believe they said they even tried flipping a Silverstone Raven 2 so that the air went sideways rather than bottom->top and saw very little if any difference.

So while heat=rises, I think "airs goes where fans push" it' is a much stronger force...

I've had a Raven 3 since it came out (https://www.silverstonetek.com/raven/products/index.php?model=RV03&area=en ) and absolutely love the case. And I think having the big fans at the bottom like that is great for easy of build/etc, but I don't really think the heat rising makes that much of a difference compared to if the case had been on its side like normal...

 

[Furious_Styles]

I believe Gamer's Nexus have flipped a few cases to see if heat=rises makes any difference, and I don't think they've really seen anything themselves. I believe they said they even tried flipping a Silverstone Raven 2 so that the air went sideways rather than bottom->top and saw very little if any difference.

So while heat=rises, I think "airs goes where fans push" it' is a much stronger force...

I've had a Raven 3 since it came out (https://www.silverstonetek.com/raven/products/index.php?model=RV03&area=en ) and absolutely love the case. And I think having the big fans at the bottom like that is great for easy of build/etc, but I don't really think the heat rising makes that much of a difference compared to if the case had been on its side like normal...

Yeah it's interesting seeing theory vs. reality. I watched a youtube vid a few weeks ago and the guy was testing the D14 or 15 and he said the best fan config he found was both on the outside blowing towards the center gap in the heatsink. Can't say I'd have predicted that.

 

[Confirmation]

I believe Gamer's Nexus have flipped a few cases to see if heat=rises makes any difference, and I don't think they've really seen anything themselves. I believe they said they even tried flipping a Silverstone Raven 2 so that the air went sideways rather than bottom->top and saw very little if any difference.

So while heat=rises, I think "airs goes where fans push" it' is a much stronger force...

I've had a Raven 3 since it came out (https://www.silverstonetek.com/raven/products/index.php?model=RV03&area=en ) and absolutely love the case. And I think having the big fans at the bottom like that is great for easy of build/etc, but I don't really think the heat rising makes that much of a difference compared to if the case had been on its side like normal...

.


thanks man, this is what i was looking for

 

[Fanatical Meat]

Slightly off topic

Does memory and SSD's put out that much heat to worry about?

 

[whm1974]

Slightly off topic

Does memory and SSD's put out that much heat to worry about?

It is my understanding that SATA SSDs produces less heat then HDDs do, but PCIe SSDs do produce a lot of heat.

Modern memory has much heat then DDR2 did.

 

[IntelUser2000]

It is my understanding that SATA SSDs produces less heat then HDDs do, but PCIe SSDs do produce a lot of heat.

Modern memory has much heat then DDR2 did.

Anything that has a "performance" tag has heatsinks nowadays. The industry seems to be aiming for 350W for datacenter CPUs, and 450W for GPU boards. We got plenty powerful power supplies after the mining craze, so its not like power delivery is an issue.

The next on the list is probably WiFi cards. At this point maybe they should just make a heatsink line that's named after form factors. So a uATX heatsink means it gets to fit over your entire uATX motherboard.

 

[whm1974]

Anything that has a "performance" tag has heatsinks nowadays. The industry seems to be aiming for 350W for datacenter CPUs, and 450W for GPU boards. We got plenty powerful power supplies after the mining craze, so its not like power delivery is an issue.

The next on the list is probably WiFi cards. At this point maybe they should just make a heatsink line that's named after form factors. So a uATX heatsink means it gets to fit over your entire uATX motherboard.

when does it stop?

 

[Fanatical Meat]

when does it stop?

When Everything is cooled by liquid nitrogen

 

[DrMrLordX]

When Everything is cooled by liquid nitrogen

LHe or bust.

 

[HisEvilness]

Heat goes up yes, you could change that with fans but it would require more work from the fans thus more noise, wise to use rising hot air for airflow just works better. But if for some reason you want to change things up the heat shouldn't stop you either way.

 

[kschendel]

It's not as simple as "heat rises". Assuming that radiation plays a negligible part at these temperatures (a reasonable assumption, I hope), you have conduction heating the air immediately around the part, and convection moving the hot air up and out. Convection generally doesn't happen as a uniform sheet, so you have hot bubbles rising and colder bubbles falling. The whole process is relatively slow. Much better to push the heated air out forcibly with fans.

Hot spots in a forced-air cooled case is an airflow design problem, not a "heat rises" problem.