Radeon HD 6850 & 6870 reviews...

[Sam_oslo]

We all know we need a cool case, for sure. But some are insisting that it's OK with hot air.
These GPU treads (both nVida nd AMD) has a tendency to become arena for claims and insisting on "rumors", I know. But lets keep it real, you don't want hot air inside your case.

[DeltZ]

bah, I don't upgrade very often at all, how long until I can double my HD4870 performance for the price I paid for it (£125), hadn't been particularly keeping up with the speculation but I was expecting these to be a bit stronger...I guess it's all about competition

[Manicdan]

open coolers give you better noise to temp ratio, if you can get the hot air out some other way.

just simple pros and cons

[Calmatory]

Are you trying to say, a hot case is the same as the cold case?
I don't know how much you know about OCing and cooling, maybe i need to explain.

Hot air from these stupid "open" coolers has sevral disadvantages:

1-Hot air gets sucked in by those big fans, and heats up you GPU too. So it's debatable if you get a cooler GPU.
2-You NB and SB are really closed by, and they get toasted.
3- Hot air will limit your OC on CPU, for sure, and would affect your RAM-OC too.


Better case-ventilation needs more fans, and it means more noise, and we should try to eliminate the source of heat for 24/7 first, and then add more fans.

1. The air is cooler than the PCB of the graphics card -> the PCB and surface mounted components release heat energy to the flowing air and become cooler. As the PCB is cooler, it radiates less heat energy to the rest of the case. Also, as the PCB is cooler, more of heat generated by GPU will be conducted to the PCB and GPU cools down.

2. As the NB and SB are close by and the graphics card PCB is cooler and the open cooler provides airflow, NB and SB cool more efficiently than with leafblower cooler design.

3. A few degrees of fluctuation in CPU temperature limit your OC? If so, I'm sure you don't pay attention to stability too much then.

4. Take a physics course.

[Sam_oslo]

open coolers give you better noise to temp ratio, if you can get the hot air out some other way.

just simple pros and cons

That can be true, if you can get the hot air out without more fans/noise. That's a big IF.

There is fine line here. That hot air will get sucked in and heat up your GPU too, so how can you be so sure that your GPU will run cooler/quieter?. What you can say for sure is: it will damage the CO-headroom for all other components, and this means you need more fans/noise.

The whole discussion started by, why can't they find a better design to eliminate disadvantages of dumping the hot air inside. So, your point is valid, but how do you get that hot air out of your case without more fans/noise? Wasn't it better if those stupid coolers could somehow dump it out by themselves?

[xdan]

1-Hot air gets sucked in by those big fans, and heats up you GPU too. So it's debatable if you get a cooler GPU.
2-You NB and SB are really closed by, and they get toasted.
3- Hot air will limit your OC on CPU, for sure, and would affect your RAM-OC too.


Better case-ventilation needs more fans, and it means more noise, and we should try to eliminate the source of heat for 24/7 first, and then add more fans.

1. You don't have hot air with 3-4 12cm fans, well pozitionated, they will suck fresh, cold air from outside the case, and exhaust the hot air.
2. P55/H55 for example don't have NB and ICH10R on 45nm is a small and cold chip. And also applies p. 1.
3., see p. 1 for Cpu, and if we speek about DDR3 memorys, DDR3 memory is very cold compared to DDR2 so no problem.
4. There are 1200-1300RPM fans with 21-26db noise, so there are very silent, silent than your VGA.

Open cooler are in VGA industry from many years, is not as they introduced them, now.
It's not about cooler , it's more about TDP card, a GTX 480(TDP 250w) open cooler produces much more heat, a GTX 460(TDP 160W) produce a little. And you not judge that even with EE cooler you dump heat in case because PCB is hoter, and it heat the air in case, to.
With an open cooler you have lower temperatures of the pcb.

Sam_oslo, you don't keep the gpu load to 100%all time, 24/7, you use him only when you play games...In idle the voltage of Gpu lower from 1.0v to 0.875v, i have with my GTX 460(Gigabyte-windforce), 35C in idle, lower temperatures than on cpu with stock cooler.
And the heat will vanish when you don't play games, so your sistem will not increase in temperature forever. Temperature into the case will lower to normal in idle periods.

Please try to keep it real, and on topic, and tell me why do you want that hot air inside the case when a good cooler could dump it outside?

For an user with so many posts, you speak nonsenses. GTX 460 isn't 8800GTX...or 9800GX2....or GTX 280...
And you are offtopic, too...this thread isn't about vga coolers. Evga has some GTX 460 external exhaust cooler models(EE), buy that to be with peace in your mind.

[Sam_oslo]

1. The air is cooler than the PCB of the graphics card -> the PCB and surface mounted components release heat energy to the flowing air and become cooler. As the PCB is cooler, it radiates less heat energy to the rest of the case. Also, as the PCB is cooler, more of heat generated by GPU will be conducted to the PCB and GPU cools down.

2. As the NB and SB are close by and the graphics card PCB is cooler and the open cooler provides airflow, NB and SB cool more efficiently than with leafblower cooler design.

3. A few degrees of fluctuation in CPU temperature limit your OC? If so, I'm sure you don't pay attention to stability too much then.

4. Take a physics course.

1. You don't have hoat air with 3-4 12cm fans, well pozitionated, they will suck fresh, cold air from outside the case, and exhaust the hot air.
2. P55/H55 for example don't have NB and ICH10R on 45nm is a small and cold chip. And also applies p. 1.
3., see p. 1 for Cpu, and if we speek about DDR3 memorys, DDR3 memory is very cold compared to DDR2 so no problem.
4. There are 1200-1300RPM fans with 21-26db noise, so there are very silent, silent than your VGA.

You guys are trying to ell me that it's OK to have hot air inside the case.?

I gotta admit you both are good at going around the point. Please try to keep it real, and on topic, and tell me why do you want that hot air inside the case when a good cooler could dump it outside?

[Calmatory]

You guys are trying to ell me that it's OK to have hot air inside the case.?

I gotta admit you both are good at going around the point. Please try to keep it real, and on topic, and tell me why do you want that hot air inside the case when a good cooler could dump it outside?

Next time try to respond to the arguments please, and don't "go around the point".

If you understood the basics of thermal physics and thermodynamics, you would understand that flowing warm air is better than cool still standing air around components, as long as the air is cooler than the components. Leafblower design delivers absolutely no airflow to the rest of the case, while open cooler provides airflow all around the case, depending on the orientation of the cooler.

It is far more important to have more airflow than cooler air for passive components like VRM, RAM, SB and NB. Thus put, open cooler isn't only more efficient at cooling, but in most cases can also provide additional cooling to other components too. Though, it is true that because of higher ambient temperature in the case, actively cooled components will most probably see higher temperatures by a few degrees. However, in practice it has no difference in OC-ability.

[Sam_oslo]

Next time try to respond to the arguments please, and don't "go around the point".

If you understood the basics of thermal physics and thermodynamics, you would understand that flowing warm air is better than cool still standing air around components, as long as the air is cooler than the components. Leafblower design delivers absolutely no airflow to the rest of the case, while open cooler provides airflow all around the case, depending on the orientation of the cooler.

It is far more important to have more airflow than cooler air for passive components like VRM, RAM, SB and NB. Thus put, open cooler isn't only more efficient at cooling, but in most cases can also provide additional cooling to other components too. Though, it is true that because of higher ambient temperature in the case, actively cooled components will most probably see higher temperatures by a few degrees. However, in practice it has no difference in OC-ability.

Actually I've studied physics in university, and you are right, airflow is good, but, (and it's big BUTT ) as you said, it has be cooler than those components, not hotter!. Look at this temps and tell me you want this to flow on your components, OK?



You want air flow, bot not from GPU or CPU-exhausts, you want fresh air flow!

[Mad Pistol]

Next time try to respond to the arguments please, and don't "go around the point".

If you understood the basics of thermal physics and thermodynamics, you would understand that flowing warm air is better than cool still standing air around components, as long as the air is cooler than the components. Leafblower design delivers absolutely no airflow to the rest of the case, while open cooler provides airflow all around the case, depending on the orientation of the cooler.

It is far more important to have more airflow than cooler air for passive components like VRM, RAM, SB and NB. Thus put, open cooler isn't only more efficient at cooling, but in most cases can also provide additional cooling to other components too. Though, it is true that because of higher ambient temperature in the case, actively cooled components will most probably see higher temperatures by a few degrees. However, in practice it has no difference in OC-ability.

I agree with this.

Here's an example.


Open case with lots of ventilation, but not much airflow

vs.

Large case with less vents, but more directional airflow.


We just upgraded my fiance's case from a RAIDMAX Sirius to an NZXT Phantom, and even though it looks like the RAIDMAX case would cool better due to nearly the entire case being vented/mesh, it didn't. The Phantom has a definite direction to all the airflow in the case, and because of that, every component is 3-8C cooler than it was in that other case.

Thermodynamics are an interesting topic, but I agree; free-standing cool air is never better than moving, slightly warmer air.





As an addendum, I think GPU heat should be exhausted out of the case as quickly as possible. There is not a single other component in the computer that generates more heat than a high-end video card. Putting that heat back into the case is ONLY a bad idea if your case doesn't have a lot of airflow. If it does have a lot of airflow, though, it shouldn't make much of a difference.

[cegras]

Actually I've studied physics in university, and you are right, airflow is good, but, (and it's big BUTT ) as you said, it has be cooler than those components, not hotter!. Look at this temps and tell me you want this to flow on your components, OK?



You want air flow, bot not from GPU or CPU-exhausts, you want fresh air flow!

I'm sure only the air in the immediate vicinity of the core area is actually around 90 C. It will diffuse (terminology?) rapidly into the surroundings. Disregarding flux of the heat into other sinks, if you simply draw a sphere radiating from the back of the GPU you see that the volume where the heat has to spread goes as the cube of the radius. The temperature should drop as the cube of the distance from the 90 C back of the PCB, as temperature is related to the amount of local heat.

So:

Passively cooled components lose heat by having it diffuse from the sink into the air. Conduction by air is terrible because air is actually not very dense at all, and heat is carried by collisions between air molecules (density of air is ~ 1.5 kg / m^3). Moving air increases the amount of collisions. Also, as hot air tends to expand and diffuse the heat throughout as much area as possible (to reach equilibrium), at the very least you're promoting convection to remove deadspots.

[Nikolasz]

Uploaded with ImageShack.us

I got pIII era case, im not noob tu buy new loud with 10 noisy fans and light material case for 50-100 euros...

Check pict.
I did dome light mod to my case i use 2 fans to blow hot air out. Thats psu and some smaller back fan, bouth are low rpm.
Then i open up slots for cards that may coold air come in to my gf9800gtx, so now as you may see on pict. Gf suck in cold air and blow it out.
Whats rest of air, it cool dowm anything else, i got heatpipes allover mainboard.
3 Hard drives as coled down same way cold air blows in, and cool them down
All working wery good and stable.
Remember hot air gows up it self
Cold air stay down

[Jowy Atreides]

bah, I don't upgrade very often at all, how long until I can double my HD4870 performance for the price I paid for it (£125), hadn't been particularly keeping up with the speculation but I was expecting these to be a bit stronger...I guess it's all about competition

£104 for a 5830 comes close.
Damaged box stock but guaranteed card

http://www.aria.co.uk/Products/Compo...roductId=42098

[Calmatory]

Actually I've studied physics in university, and you are right, airflow is good, but, (and it's big BUTT ) as you said, it has be cooler than those components, not hotter!. Look at this temps and tell me you want this to flow on your components, OK?

You want air flow, bot not from GPU or CPU-exhausts, you want fresh air flow!

Thats GPU temperature, and if the GPU was the ONLY source of heat energy to that array of fins in the cooler, the exhaust air could NEVER reach the GPU temperature. Hence the exhaust air is far from being as warm as the GPU.

Actually, it would be a miracle if the exhaust air was even near the temperature of the cooler fins. Which itself should be as warm as the base of the heatsink, which itself should be as warm as the heatspreader, which itself should be as warm as the GPU die etc. GPU temperature has very little to do with exhaust air temperature, especially if talking about open GPU coolers.

Of course there's possible problem with graphics card VRM cooling depending on the orientation of the cooler etc.

[RSC]

Stop with the off topic guys... Come on.

[Calmatory]

Stop with the off topic guys... Come on.

And pull something new and interesting out from our hats for you to enjoy? If there's someone with something contributive to say, he will do so regardless of what other people discuss meanwhile.

Stop with your offtopic and make me interested!

[BababooeyHTJ]

My northbridge temps went up a few degrees after I slapped an HR-03GTX onto my GTX280. This was in a Antec 1200 loaded with 1300rpm Yate Loons.

It was worth it for the drop in temps on the gpu itself but I don't know that would be the case in a case without really good airflow.

Thats just what I saw.

That said I want a non reference 6870 with an open cooler.

[Manicdan]

i just built a case in a briefcase (see it under the water cooling worklogs), and it has ONLY 1 fan in and 1 fan out. the cpu gets the coldest air since AMD chips are very touchy to temps, the gpu gets the spare cpu heat, and the psu pushes out the rest to where the hard drives are and the exhaust. my OC was not in any way limited because of this setup. AND if it open the top which is right above the GPU, the temps GET WORSE. cause theres no pressure remaining to force air out. when running furmark and prime my temps are 38C on cpu, and 82c gpu after a half hour. yes other things got hot, but that wasnt affecting anything i needed since 25C vs 40C isnt hurting the NB, RAM, or whatever else can be OCed.

[blindbox]

And pull something new and interesting out from our hats for you to enjoy? If there's someone with something contributive to say, he will do so regardless of what other people discuss meanwhile.

Stop with your offtopic and make me interested!

This post made me laugh, yet I truly agree!

[oohms]

The air coming from those open sink graphics cards mostly gets blocked by its own PCB, so very little of that actually blows onto the cpu/NB, and the SB rarely needs extra airflow. It just lingers around and heats the case up

There is no reason not to have external exhaust with graphics cards - results in cooler case temps, as well as cooler intake temps for the graphics card itself

[DedEmbryonicCe1]

Some people like myself go without a case entirely or use open air test benches for 24/7 use. As such the top->down air flow coolers are superior to the ones that try and force hot air down a narrow channel and through some tight slots at the PCI bracket. It is up to the consumer to purchase a card with the right cooler for their situation and blanket generalizations don't do anyone any good when everyone benefits from having their choice of either design from manufacturers.

[Nintendork]

XFX 3D ad:

http://www.youtube.com/watch?v=rG4klFf2cqk

[ripken204]

XFX 3D ad:

http://www.youtube.com/watch?v=rG4klFf2cqk

so 3D will make you look like a pansy and piss your pants haha

[STEvil]

The air coming from those open sink graphics cards mostly gets blocked by its own PCB, so very little of that actually blows onto the cpu/NB, and the SB rarely needs extra airflow. It just lingers around and heats the case up

There is no reason not to have external exhaust with graphics cards - results in cooler case temps, as well as cooler intake temps for the graphics card itself

If its "lingering" in the case then you have an improperly vented case.

[hawk999]

6870 meant to replace 5770? is that true cause i watch reviews in youtube by motherboardsorg