AMD "Barts" XT card pixellized

Sickamore

Senior member
Aug 10, 2010
368
0
0
http://www.fudzilla.com/graphics/graphics/graphics/amd-barts-xt-card-pixellized

amd_barts_1.jpg

amd_barts_2.jpg





You can read more about it here.
 

Grooveriding

Diamond Member
Dec 25, 2008
9,147
1,330
126
Looking at the size of the bottom component for the dual DVI connectors, it's a safe bet this card will come with a dual slot cooler when it is released to market.

Also looks like even their new mid range are getting the full gamut of outputs with 2 mini-display port, and hdmi and two DVI.
 

Lonyo

Lifer
Aug 10, 2002
21,938
6
81
More pressing is the dual PCI-e power connectors, suggesting above or close to 150w power consumption.
 

Daedalus685

Golden Member
Nov 12, 2009
1,386
1
0
More pressing is the dual PCI-e power connectors, suggesting above or close to 150w power consumption.

Aye, but it is quite common for an engineering sample to have much more possible power than required no? Didn't the first juniper cards demonstrating eyefinity also have dual 6pin?

Everything about this kind of screams "sample card" to me. I've heard ATI is moving towards vapour chambers though (S/A forum rumours) for most of the mid end and up. This does not look like a vapour chamber to me (if it is it could explain the "crappy" look of it) but it is hard to tell.
 

Grooveriding

Diamond Member
Dec 25, 2008
9,147
1,330
126
Aye, but it is quite common for an engineering sample to have much more possible power than required no? Didn't the first juniper cards demonstrating eyefinity also have dual 6pin?

Everything about this kind of screams "sample card" to me. I've heard ATI is moving towards vapour chambers though (S/A forum rumours) for most of the mid end and up. This does not look like a vapour chamber to me (if it is it could explain the "crappy" look of it) but it is hard to tell.

I'm going to wager it comes to market in the dual 6pin configuration. It may even require as much or a little more power than a GTX 460 if the early leaks and rumors are true about it being faster than the 460.
 

Skurge

Diamond Member
Aug 17, 2009
5,195
1
71
I'm going to wager it comes to market in the dual 6pin configuration. It may even require as much or a little more power than a GTX 460 if the early leaks and rumors are true about it being faster than the 460.

A 5850 is faster than a 460 but uses a bit less power.

They made the new core more power efficient so it might still use less power than a 460 and still be faster.
 

bryanW1995

Lifer
May 22, 2007
11,144
32
91
as long as the performance is there then 2x6 pins is fine. all the #'s point to a card that will be similar to or faster than 5870. barts pro will probably be a better card for low power/midrange, barts xt is looking like it might creep up into the high end this round.
 

lopri

Elite Member
Jul 27, 2002
13,314
690
126
That is a weird sample. Why would they locate VRMs on the opposite side of power connectors? :confused:
 

faxon

Platinum Member
May 23, 2008
2,109
1
81
Looks like that cooler was designed by coolermaster like a lot of AMDs past stock CPU coolers. Check the fan blade design, only CM uses those right now afaik. Looks kind of wimpy though compared to the stock on the 5850, maybe the count about a vapor chamber design was spot on? Or is it possible they just skimped on cooling cause it's an ES
 

StrangerGuy

Diamond Member
May 9, 2004
8,443
124
106
Looks like that cooler was designed by coolermaster like a lot of AMDs past stock CPU coolers. Check the fan blade design, only CM uses those right now afaik. Looks kind of wimpy though compared to the stock on the 5850, maybe the count about a vapor chamber design was spot on? Or is it possible they just skimped on cooling cause it's an ES

The fan is pretty big. I like my fans as big as possible. They move a lot of air without a high-pitch whine the smaller ones do.
 

Daedalus685

Golden Member
Nov 12, 2009
1,386
1
0
That's actually kinda exciting to me. I'm tired of GPUs spewing their heat out into my case. I want it all exhausted out the back.

Aye.

Perhaps they figured cooling the VRM after the GPU rather than before would result in better overall cooling. The little buggers get very hot, but unlike the GPU there is a bit more acceptability to them being as such, so a higher ambient (from already warmed GPU air) is not going to hurt as much.
 

bryanW1995

Lifer
May 22, 2007
11,144
32
91
but if they're already so hot then cooling them is more difficult already. best benefit I can see is that it is close to the exit and thus less of the heat pollutes your case.
 

Grooveriding

Diamond Member
Dec 25, 2008
9,147
1,330
126
Looks like that cooler was designed by coolermaster like a lot of AMDs past stock CPU coolers. Check the fan blade design, only CM uses those right now afaik. Looks kind of wimpy though compared to the stock on the 5850, maybe the count about a vapor chamber design was spot on? Or is it possible they just skimped on cooling cause it's an ES

Have AMD used a reference design like that before ? I know I've seen that on custom cards, I can't remember seeing it on a reference card though.
 

Daedalus685

Golden Member
Nov 12, 2009
1,386
1
0
but if they're already so hot then cooling them is more difficult already. best benefit I can see is that it is close to the exit and thus less of the heat pollutes your case.

The air in most dual slot configurations enters at the far end from the ports and exits through the vent above the ports. The cooling effectiveness is related to the temperature of this air as it passes over any given heat sink. Since the VRMs can accommodate a higher temperature their ambients can be higher for any given cooling system. So you might save a degree or two (depending on cooling system) by ensuring the coolest air arrives at the more temperature sensitive GPU.

The position does not affect heat dumped into the case if they use the same cooler as last time around. All of the air should be flowing from one end to the other with the only case dumped heat from radiative losses. Something that should not be all that different. However, if they use a central fan like in this picture the location pictured is the only location for the VRMS that would allow air to carry the heat out of the case. The other side would push it back in.
 

Borealis7

Platinum Member
Oct 19, 2006
2,901
205
106
i like the circular heat-sink design as i have 2x120mm fans on the side panel of my case just above the video card. keeps it nice and cool. gimme more o' those.
 

nitromullet

Diamond Member
Jan 7, 2004
9,031
36
91
Have AMD used a reference design like that before ? I know I've seen that on custom cards, I can't remember seeing it on a reference card though.

I doubt this will be the reference cooler. You can actually see where the outline of the stock cooler is intended to be because it's been silk screened onto the PCB.
 
Dec 30, 2004
12,553
2
76
The air in most dual slot configurations enters at the far end from the ports and exits through the vent above the ports. The cooling effectiveness is related to the temperature of this air as it passes over any given heat sink. Since the VRMs can accommodate a higher temperature their ambients can be higher for any given cooling system. So you might save a degree or two (depending on cooling system) by ensuring the coolest air arrives at the more temperature sensitive GPU.

The position does not affect heat dumped into the case if they use the same cooler as last time around. All of the air should be flowing from one end to the other with the only case dumped heat from radiative losses. Something that should not be all that different. However, if they use a central fan like in this picture the location pictured is the only location for the VRMS that would allow air to carry the heat out of the case. The other side would push it back in.

I completely agree. They can withstand 125C...cool them second....give the GPU the cooler air.
 

Rubycon

Madame President
Aug 10, 2005
17,768
485
126
That is a weird sample. Why would they locate VRMs on the opposite side of power connectors? :confused:

VRM closer to the GPU results in cleaner power and less influence with longer runs, etc. ?