AMD R9 390X Coming With Cooler Master Liquid Cooler + Estimated Performance

[Fetzie]

I don't like aio's and I CBF to do a custom loop so this is dissapointing, I hope this doesn't become a reg thing for high end GPUs

 

There'll be air-cooled 3rd party cards.

[HippY]

 

 

Also, a 20 dB(A) difference is 100 times as loud, not 4 times as loud. It's a base-10 logarithm, not a base-2 logarithm. 20 dB(A) = 2 B(A) = 10^2 = 100

so you are basically saying that the R9 290X is as loud as 100x 970?

 

are you sure you no logs well enough?

[Sakkura]

so you are basically saying that the R9 290X is as loud as 100x 970?

 

are you sure you no logs well enough?

 

Not quite. But 10^2 really does equal 100. Are you saying it doesn't?

 

Also, as I said, I reject their measurements as unreliable.

[carolkarine]

The problem with the devil 13 is that it performed poorly compared to the reference design. I'm hoping that the AIB's introduce really high-end air coolers that would actually be competitive with the reference design.

Sorry if this is ridiculously late or if someone already said it, but the 295x2 had a tdp of 500w, and that was a low estimate. It could pull a good deal more. So the only cooler on the market which even tries to claim it can dissipate that much heat is gigabyte. However, the 390x is rumored to have a 300w tdp. Nothing to scoff at, I'm sure, but it's not a 600w monster. Hell, some mildly oc'ed 290x's reach that regularly. I think we'll definitely see air coolers for it, just because even if the reference board is water cooled, it doesn't mean air coolers won't be able to handle the heat.

I do expect the AiB cards to be CHEAPER, because I would imagine an AIO plus heatsink for other components would be more expensive to manufacture than a single heatsink, even if it is massive.

[Sakkura]

Sorry if this is ridiculously late or if someone already said it, but the 295x2 had a tdp of 500w, and that was a low estimate.

 

Actually that was on the high side. Power consumption in games is in the 430W range, 450W in GPGPU.

 

http://www.tomshardware.com/reviews/radeon-r9-295x2-review-benchmark-performance,3799-16.html

 

http://www.techpowerup.com/reviews/AMD/R9_295_X2/22.html

[carolkarine]

Actually that was on the high side. Power consumption in games is in the 430W range, 450W in GPGPU.

http://www.tomshardware.com/reviews/radeon-r9-295x2-review-benchmark-performance,3799-16.html

http://www.techpowerup.com/reviews/AMD/R9_295_X2/22.html

Fair enough. There's still a BIG difference between 300w and 450w. And gigabyte is STILL the only manufacturer that claims that kind of dissipation.

[JAKEBAB]

There'll be air-cooled 3rd party cards.

Im 90% sure there was only 1 air cooled 295x2 and it was like a 3 slot card

[mr moose]

SPL is what dB is used to measure. But that's where 10 dB means 10 times as loud. You want to convert that to subjective loudness, which there is no objective way to do. Phons and sones are units attempting to quantify perceived loudness, but they are not objective measures.

 

When dB is used as the measurement of SPL. it is used as a reference scale. Your understanding of log scales is moot, because we are talking about SPL. just because the scale is log doesn't change the relevant points of reference on that scale.  The SPL is a very measurable physical property.  When someone refers to dB with regard to loudness or SPL it is a very well defined parameter.  To argue 20db = 100x louder is ludicrous.  10dB is perceived as 4x as loud.  So unless you can find something to support your claims please stop making thing ups.

[Fetzie]

Im 90% sure there was only 1 air cooled 295x2 and it was like a 3 slot card

 

Yes, but this is half the TDP of that card. Heck, there were plenty of air coolers for the GTX 480 and that ran hotter than this will.

[Sakkura]

When dB is used as the measurement of SPL. it is used as a reference scale. Your understanding of log scales is moot, because we are talking about SPL. just because the scale is log doesn't change the relevant points of reference on that scale.  The SPL is a very measurable physical property.  When someone refers to dB with regard to loudness or SPL it is a very well defined parameter.  To argue 20db = 100x louder is ludicrous.  10dB is perceived as 4x as loud.  So unless you can find something to support your claims please stop making thing ups.

 

SPL is a physical property, loudness is not. You're the one making unsubstantiated claims about loudness, I've already demonstrated that 20 dB means a 100-fold difference. That's literally what 20 dB means. I don't know why you can't accept simple facts.

[JAKEBAB]

Yes, but this is half the TDP of that card. Heck, there were plenty of air coolers for the GTX 480 and that ran hotter than this will.

Oh i heard it's TDP was going to be something like 400w (which doesn't sound real at all) but it was a rumor (i know rumor) for quite awhile so im assuming it's alot less now? When they ship the card with reference aio i just assume it's going to be a hot and power hungry card because why the hell would i need an aio if it wasn't? but whatever it's more than likely going to be air coolers if what your saying is true so thats good to know.

[Fetzie]

Oh i heard it's TDP was going to be something like 400w (which doesn't sound real at all) but it was a rumor (i know rumor) for quite awhile so im assuming it's alot less now? When they ship the card with reference aio i just assume it's going to be a hot and power hungry card because why the hell would i need an aio if it wasn't? but whatever it's more than likely going to be air coolers if what your saying is true so thats good to know.

 

300W, and AMD quotes the maximum possible TDP instead of a "typical use" scenario.

[mr moose]

SPL is a physical property, loudness is not. You're the one making unsubstantiated claims about loudness, I've already demonstrated that 20 dB means a 100-fold difference. That's literally what 20 dB means. I don't know why you can't accept simple facts.

 

Stop posting rubbish and get yourself educated:

 

https://www.osha.gov/dts/osta/otm/noise/health_effects/soundpressure_aweighted.html

http://trace.wisc.edu/docs/2004-About-dB/

 

A weighted  measurements are used specifically for calculating the relative volume/loudness of a device to the average listener.  There is a direct  correlation to perceived volume and the measured dB. 

 

You can argue loudness is subjective and all that, but at the end of the day for static noise (like a fan and not music which is dynamic), A weighted SPL correlates to perception.

 

So again, unless you can post something that disputes the myriad of knowledge amassed and used by professionals around the world on the subject stop telling people that 20dB is 100 times louder. it is not.

[Ashley]

I don't like aio's and I CBF to do a custom loop so this is dissapointing, I hope this doesn't become a reg thing for high end GPUs

What if someone releases a card with expandable water cooling?

Basically an AIO would be installed by default but would be easily detached and hooked into a custom loop. I believe the Poseidon cards are a similar idea to this.

[JAKEBAB]

What if someone releases a card with expandable water cooling?

Basically an AIO would be installed by default but would be easily detached and hooked into a custom loop. I believe the Poseidon cards are a similar idea to this.

Can the poseidon cards run without any type of watercooling just on air? because i know they have fans but i don't know if there just to make it even cooler, do you know? Maybe that could be somewhat of a solution if they didn't have the tubing sticking out so much or have some sort of cover in the shroud.

 

But the whole point of doing the water cooling anyway is because a fan can't handle it  If the fan can handle it there is no point in giving the consumer basically a free water block when they don't need to. They would probs go back to what we have now anyway, and the people who want to water cool there cards can and the people who dont, dont have to

[Bsmith]

300W, and AMD quotes the maximum possible TDP instead of a "typical use" scenario.

 

yeah, but waht do they see as "typical use"?

it's a high end card so i think it would be gaming, screen capture, rendering etc, which makes sense with the AIO included which also covers enthousiast zone.

But its possible they see office systems, show systems and quiet systems(since the AIO is quiter) with minimal load on the GPU,

 

man, i wish AMD would shove some real info/numbers out there, to many rumours and possibilites at the moment.

 

 

Can the poseidon cards run without any type of watercooling just on air? because i know they have fans but i don't know if there just to make it even cooler, do you know? Maybe that could be somewhat of a solution if they didn't have the tubing sticking out so much or have some sort of cover in the shroud.

 

the poseidon's are hybrid cards, which means they can run on air, water or a combination of both.

of course this isn't really optimal since both air and water use diffrent ways of cooling (air= heatpipe, water=direct contact) so you get a kind of fusion of both, which doesnt work out very well most of the time, altough IIRC the latest poseidon cards had a good solution, which made both ways viable to do.

 

 

But the whole point of doing the water cooling anyway is because a fan can't handle it  If the fan can handle it there is no point in giving the consumer basically a free water block when they don't need to. They would probs go back to what we have now anyway, and the people who want to water cool there cards can and the people who dont, dont have to

 

this is also what bother me, i wonder how it would go with prices, 3rd party air coolers basicly could be cheaper(and better) then the reference design since they don't have the more expensive liquid cooling.

and yeah, most ppl that WC do it custom by their own, i admit you get already a nice block now to start with, but connecting the tubing and pump inside it would be a problem though for custom watercooling.

[Sakkura]

Stop posting rubbish and get yourself educated:

 

https://www.osha.gov/dts/osta/otm/noise/health_effects/soundpressure_aweighted.html

http://trace.wisc.edu/docs/2004-About-dB/

 

A weighted  measurements are used specifically for calculating the relative volume/loudness of a device to the average listener.  There is a direct  correlation to perceived volume and the measured dB. 

 

You can argue loudness is subjective and all that, but at the end of the day for static noise (like a fan and not music which is dynamic), A weighted SPL correlates to perception.

 

So again, unless you can post something that disputes the myriad of knowledge amassed and used by professionals around the world on the subject stop telling people that 20dB is 100 times louder. it is not.

 

You are still not posting any proof for your claim that 10^2 does not equal 100, nor that a 20 dB difference equals the loudness difference you claim.

 

A 20 dB difference is not a 100-fold increase in perceived loudness, but it is a 100-fold physical difference. How our ears and brains perceive that difference is complicated and non-uniform.

 

A weighted SPL does not correspond to perceived loudness, because A weighting is merely an approximate empirical adjustment to account for the frequency response of the human auditory system. But it does not account for other factors in the auditory system.

[mr moose]

You are still not posting any proof for your claim that 10^2 does not equal 100, nor that a 20 dB difference equals the loudness difference you claim.

 

A 20 dB difference is not a 100-fold increase in perceived loudness, but it is a 100-fold physical difference. How our ears and brains perceive that difference is complicated and non-uniform.

 

A weighted SPL does not correspond to perceived loudness, because A weighting is merely an approximate empirical adjustment to account for the frequency response of the human auditory system. But it does not account for other factors in the auditory system.

 

please? 

 

You have totally missed the point of taking these measurements in the first place?  No one cares about the log scale, it is and has always been about perceived volume/loudness. That is what the A weighting is for and that is why it is measured as such.   We are talking about how much noise the cards make. The fact that you can't apply that knowledge contextually is proof you don't understand sound measurement. 

 

Again Again;  if you want to tell me I am wrong then post proof, I have already posted links to two separate sites that explain why it is 4x louder, you where the one who tried to insinuate it was 100x.

[Sakkura]

please? 

 

You have totally missed the point of taking these measurements in the first place?  No one cares about the log scale, it is and has always been about perceived volume/loudness. That is what the A weighting is for and that is why it is measured as such.   We are talking about how much noise the cards make. The fact that you can't apply that knowledge contextually is proof you don't understand sound measurement. 

 

Again Again;  if you want to tell me I am wrong then post proof, I have already posted links to two separate sites that explain why it is 4x louder, you where the one who tried to insinuate it was 100x.

 

But your own sources disagree with the claims you are making. You're claiming 10 dB is exactly a 4x difference in loudness, while your sources explain that 20 dB (not 10 dB!) is only very roughly equal to a 4x difference in perceived loudness. And you were also making errors when adding up noise from multiple sources. Even if we ignored interference, it takes a lot more than 4x as many noise generators to equal noise generators that are 20 dB louder. Because the physical difference is much bigger - a 20 dB difference is a 100x difference in power, 10x difference in pressure.

[Materi]

i truly love how deraild, childish and truly uninformative this thread has gotten... its even worse than the OC.net thread on the same topic. but people need too realize that there are close to no information about the 300 series from AMD. all the FACTS that WCCF are using in their artical is a shipping record and that shipping record stats "Cooler Master Heatsink" not AIO or water cooling.... we have abselutly none information on how many SPs that are on the die, or how many CUs, Rops, vram or anything else. we cant even be sure that AMD will be using HBM as there are zero evidence. the only thing that is know is that its coming ann graphics card, soon. and here is a link to all the roumors, facts and suchs on videocardz

http://videocardz.com/54727/amd-radeon-r9-300-series-rumors

[mr moose]

But your own sources disagree with the claims you are making. You're claiming 10 dB is exactly a 4x difference in loudness, while your sources explain that 20 dB (not 10 dB!) is only very roughly equal to a 4x difference in perceived loudness. And you were also making errors when adding up noise from multiple sources. Even if we ignored interference, it takes a lot more than 4x as many noise generators to equal noise generators that are 20 dB louder. Because the physical difference is much bigger - a 20 dB difference is a 100x difference in power, 10x difference in pressure.

 

Clearly you didn't read them. 

from the article I linked:

Perception of Loudness   (20dB = 4x)

Interestingly, our perception of loudness is not the same as sound pressure level.   Although the actual formulae

is somewhat complex,  as a rough rule of thumb, an increase of 10db SPL is perceived to be approximately twice as loud. 

    Thus a 20 Db gain would seem to be about 4 times as loud.  

    And a 40 Db gain would seem to be about 16 times as loud.

 

 

this isn't exactly rocket science, there is nothing more to say.

[Sakkura]

Clearly you didn't read them. 

from the article I linked:

 

this isn't exactly rocket science, there is nothing more to say.

 

Yes, that quote directly contradicts your claim that 10 dB = 4 times as loud.

[mr moose]

Yes, that quote directly contradicts your claim that 10 dB = 4 times as loud.

 

No it doesn't. 

[Sakkura]

No it doesn't. 

 

It says 20 dB is very roughly equivalent to 4 times as loud. You said 10 dB is exactly equal to 4 times as loud. Not only are you way off with the number, you're also claiming a precision that does not exist here.

[mr moose]

It says 20 dB is very roughly equivalent to 4 times as loud. You said 10 dB is exactly equal to 4 times as loud. Not only are you way off with the number, you're also claiming a precision that does not exist here.

 

you must really hate being wrong.  Even to the point of adding adverbs and then arguing semantics you are still trying to avoid admitting this is wrong:

 

Sakkura, on 21 Feb 2015 - 1:32 PM, said:

 

    

    Also, a 20 dB(A) difference is 100 times as loud, not 4 times as loud. It's a base-10 logarithm, not a base-2 logarithm. 20 dB(A) = 2 B(A) = 10^2 = 100