Wednesday, 14th, probably midnight, maybe 8AM PST.Joe DeFuria said:
NV40 Test samples, pics inside
- Thread starter surfhurleydude
- Start date
Joe DeFuria
Legend
The Baron said:
Cool, I knew someone would spill it if I got it wrong.
It's more of a guess than "Hey, that's when it lifts," but I know for sure that it is the 14th. The times are totally made up, though.Joe DeFuria said:The Baron said:
Cool, I knew someone would spill it if I got it wrong.
Fred said:
Just be sure to wear your Anand BS blocking glasses, the glare from his bias is rather blinding.
Lecram25 said:Told y'all there was a heatpipe
Ummm no one contended that. Anyone with 2 eyes can see them in the pic and given the fact that heat output will most likely be greater than the NV30, what else are they going to use to move heat? Pelt? H2O?
I am really curoius, but does any wonder if there will be any problems with that card falling out of the AGP slot? 
It seems to me with that amount of weight on the card and with the extreme amount of heat generated that the card might possible flex or fall out of the AGP slot even with the conventional plastic brackets that are supposed to hold the cards in.
Raystream
It seems to me with that amount of weight on the card and with the extreme amount of heat generated that the card might possible flex or fall out of the AGP slot even with the conventional plastic brackets that are supposed to hold the cards in.Raystream
I wonder how the RAM fares being hooked up to the same fin/pipe assembly as the core. GDDR3 might be cooler running, but heating up the environment arond the chip by bringing it into contact with hot air and metal won't help it dissipate whatever reduced amount of heat it makes.
Perhaps the RAM wouldn't care too much whilst running at stock, though if there was any benefit to cooling the RAM, it is obviated by heating up the ramsinks first with the core.
Is it possible that the ramsink and heatpipe assembly aren't there for the RAM at all, but instead are necessary to supply even more surface area for the core's heat dissipation needs within the space constraints of a video card (even if 2 slots are used)? If the RAM could shrug off the heat output, then tacking on the RAM cooling assembly could help cool the core down.
Perhaps the RAM wouldn't care too much whilst running at stock, though if there was any benefit to cooling the RAM, it is obviated by heating up the ramsinks first with the core.
Is it possible that the ramsink and heatpipe assembly aren't there for the RAM at all, but instead are necessary to supply even more surface area for the core's heat dissipation needs within the space constraints of a video card (even if 2 slots are used)? If the RAM could shrug off the heat output, then tacking on the RAM cooling assembly could help cool the core down.
Speaking of gamma adjusted images... anyone noticed that the heatsink does not continue below the blower? You can clearly see the green of the PCB. This doesn't strike me as a particularly effective design... or does this contribute to airflow so much that it's worth the sacrificed heatsink surface?
hoom
Veteran
NV have gone & gotten the centrifugal pump completely wrong again :? Many of the same mistakes
- too small blades & too far off axis = needs high rpm to get much draw
- blade pitch is backwards = may actually work better if the fan rotates in opposite direction
- blade concavity is backwards = would almost work better for sucking air through the heatsink first, then pushing out the fan hole
- blade pitch/concavity, fan handedness & housing handedness are mutually exclusive to optimum operation
The right way:
http://www.cheresources.com/centrifugalpumps2.shtml
Notes:
- this is the basic configuration of nearly all centrifugal pumps ever produced
- the discharge is the 'wide bit' (if nv40 sample is clockwise fan, then it is discharging through the narrow bit = wrong)
- the blade spin is anti clockwise (as I assume for nv40 sample pic)
- the pitch direction of the blades
- the concavity direction of the blades
- marketing pic of single slot nv40 has a nearly correct (clockwise rotating -not important except for clarity) fan -only has the concavity reversed
In a mixed flow design, the fan produces a mixture of axial and centrifugal flow. (note the ATI fans actually have a fair bit of room underneath them for air to be pushed into)
Though, only the 9800np/p & 9600xt actually have the right fan to properly be mixed, with the other cards having typical axial flow fans which pull the air in towards the axis rather than pushing it outwards, thus leaving simple over-pressure in the centre to provide the outwards component.
A 9800xt with an enlarged 9800np/p type fan would have been a very good mixed flow air pump (could have been even better if it had an extra slot width for the axial component?).
StealthHawk
Regular
radar1200gs said:
You realize that most people don't seem to understand the difference between a review and a preview I hope. You don't have to look far to see people in forums all over the place calling previews reviews.
radar1200gs
Regular
StealthHawk said:
Unfortunately, I understand only too well. I'm sick of sparring with the mindless morons. Let them think whatever they want to think.
Richthofen
Newcomer
Joe DeFuria said:
well that's a problem both IHVs will face. Both have the same access to the latest memory technology. Maybe both IHVs will start with some modest clocked cards and will refresh them a littlebit later when faster memory will be available.
If they overclock the memory - well no problem as all as long as they guarantee that it will work flawless. And that is what they have to do in such a scenario. I see no problem there.
I was like, when I saw that pic on the first page... DAMN!
Don't mean to sound like a negative bitch or anything, but what the HELL is going on at Nvidia these days? I mean, they USED to call fast computers 'big iron', are they on some kind of retro trip over there at NV HQ? And what's up with that fan anyway? Like already pointed out above, it's completely backwards, blades pointing in the wrong direction and the flow channel positioned incorrectly based on the impeller's direction of rotation (assuming it's meant to blow that is, if it's supposed to suck then it's merely the flow channel that is positioned incorrectly). What a botchjob!
Ok, so it's only a prototype/review sample/whatever. If retail cards are delivered with better coolers this complaint goes right out the window. I sure hope though the chip's performance makes up for this weird-ass cooler, and I think it does. Still wondering if their AA will be any damn good, but I am still excited about NV40...
The geek in me will be disappointed in ATi if they deliver a product with on-par or better performance and superior AA whilst dissipating less heat, yet lack shaders v3.0. However, I try not to listen to him too much. Don't know what I'll think of R420 if NV40 turns out to be hard to beat regarding performance. If it has both that and better shaders going for it...
This week will be very interesting!
when I saw that pic on the first page... DAMN!Don't mean to sound like a negative bitch or anything, but what the HELL is going on at Nvidia these days? I mean, they USED to call fast computers 'big iron', are they on some kind of retro trip over there at NV HQ? And what's up with that fan anyway? Like already pointed out above, it's completely backwards, blades pointing in the wrong direction and the flow channel positioned incorrectly based on the impeller's direction of rotation (assuming it's meant to blow that is, if it's supposed to suck then it's merely the flow channel that is positioned incorrectly). What a botchjob!
Ok, so it's only a prototype/review sample/whatever. If retail cards are delivered with better coolers this complaint goes right out the window. I sure hope though the chip's performance makes up for this weird-ass cooler, and I think it does. Still wondering if their AA will be any damn good, but I am still excited about NV40...
The geek in me will be disappointed in ATi if they deliver a product with on-par or better performance and superior AA whilst dissipating less heat, yet lack shaders v3.0. However, I try not to listen to him too much.
Don't know what I'll think of R420 if NV40 turns out to be hard to beat regarding performance. If it has both that and better shaders going for it...This week will be very interesting!