Cooling Revolution.

[Rodéric]

You forgot to factor in the NGC pads & memory cards ports on the left flank...
And the url of the first attachment in kinda borked, use http://www.beyond3d.com/forum/attachment.php?attachmentid=79&d=1133333198 instead.

 

[PC-Engine]

You forgot to factor in the NGC pads & memory cards ports on the left flank...
And the url of the first attachment in kinda borked, use http://www.beyond3d.com/forum/attachment.php?attachmentid=79&d=1133333198 instead.

Fixed.:smile:

As for the GCN ports just shift the heatsink to the right and downward since I left some room on the right/bottom for who knows what reason.

Also don't forget to shift the CPU/GPU toward the center too.

 

[Teasy]

PC-Engine, Revolution will be a bit wider then that. The 5" slim DVD drive will be the same width as the DVD opening at the front. So there'll then be about half to three quarters of an inch of case at each side of the drive. I don't want to nit pick or anything, just FYI

 

[PC-Engine]

PC-Engine, Revolution will be a bit wider then that. The 5" slim DVD drive will be the same width as the DVD opening at the front. So there'll then be about half to three quarters of an inch of case at each side of the drive. I don't want to nit pick or anything, just FYI

Heh yeah I realize that. I made those drawings without any measurements whatsoever. They were completely done by eye and was done in less than 15 minutes. :smile:

The round circle representing the DVD disc should be a little smaller relative to the drive housing.

Oh btw I used a picture from Anandtech where he uses a AA battery as reference and it turns out the heatsink over the GPU is ~ 3.75" x 2.75" which is slightly smaller than I had estimated, however I got the 3.75" exactly right. It's basically about the size of two bare 2.5" HDDs stacked on top one another..very small indeed.

 

[SlmDnk]

I tried a couple of them...(including Ageia's site). No luck.

If you believe that's the whole story...

The "28 watts" comes from this Andy Keane interview: http://techgage.com/review.php?id=2038&page=3

Andy Keane: The total power consumption for the board is around 28 watts. The fans and cooling will be appropriate to that, and therefore we don't expect power and cooling to cause any related issues.

Before that it was 20 watts.

 

[pc999]

Thank you very much for the info, do you know why the jump from 20W to 28W?

BTW

normal VMX is 4 milion transistors component that can do 4 op/cycle, take for example four of them with a dedicated L2 Cache and you will have 40GFLOPS extra in a system that runs at 2.5Ghz.

I they really are using 1T-SRAM-Q would they really need L2 at all, or at least a lot (in trassistores counts 512kb of L2 is already ~33M (I think but I am not sure), so it is many).

 

[Urian]

I am starting to believe that perhaps the Revolution is a mobile version of the Xbox360.

 

[Teasy]

Thank you very much for the info, do you know why the jump from 20W to 28W?

28 watts is the power consumption of the whole add in board according to those quotes. So 20 watts is probably the power consumption of the PPU chip alone.

 

[darkblu]

on a far parallel, it might be curious to see the mac mini's consumption under different conditions - i found some data posted on this site

I plugged the power brick of my Mac Mini into a simple integrating power meter. Here is what I measured:

Off 0-2 Watts
Booting 30-40 Watts
Idle 25 Watts
Sleep 3-5 Watts, almost always 3 Watts

The power brick is rated at 85 Watts output. I have yet to measure power consumption during a compute-intensive task such as DVD playback.

I think the 40 Watt max was during hard disk and DVD spin-up at boot time. Idle means that the disk is spinning, booted, logged in, at the Finder with no user input.

 

[pc999]

Ok thanks.

 

[OtakingGX]

Any rough estimates on what internal space would allow for a heat sink inside Revolution's case? I'm estimating 10cm*6cm*2.4cm for a heat sink covering Hollywood and Broadway. I'm going to put together some geometry in Pro/E and import it into Fluent and see if I can find a peak temperature given some estimates for heat generation.