RTX 3080 image (maybe?)

[StDragon]

2 minutes ago, Vishera said:

That's what makes it fluid dynamics and not hydrodynamics or aerodynamics...

You used the wrong term...

FYI, air (gas) becomes fluid at high speed. In fact, fluid dynamics applies to gas flow in general. Just ask GE, they design turbines.

[Vishera]

Just now, StDragon said:

FYI, air (gas) becomes fluid at high speed. In fact, fluid dynamics applies to gas flow in general. Just ask GE, they design turbines.

My point is that hydrodynamics and aerodynamics are fluid dynamics,but hydrodynamics is not aerodynamics.

[Bombastinator]

1 hour ago, suicidalfranco said:

the reason why i'll never buy reference 

 

Ironically this is the exact same reason I always at least consider reference.  I take your point though.

[Bombastinator]

19 minutes ago, Vishera said:

My point is that hydrodynamics and aerodynamics are fluid dynamics,but hydrodynamics is not aerodynamics.

Semanitcs.  The math is identical.  You’re not wrong though.  This gets into whether penguins fly.  I think they do.  Just not in air.

[tor6770]

Hey guys, I have made an RTX render animation on Unreal Engine 4 of my RTX3080 model

 

[StDragon]

7 minutes ago, tor6770 said:

Hey guys, I have made an RTX render animation on Unreal Engine 4 of my RTX3080 model

Fantastic! Had this video been available prior, I would have voted differently on the design

[Vishera]

13 minutes ago, tor6770 said:

Hey guys, I have made an RTX render animation on Unreal Engine 4 of my RTX3080 model

 

I think that NVIDIA uses different vent design on the IO shield,other than that it looks fairly accurate.

Good Job 

 

Example:

[tor6770]

11 minutes ago, Vishera said:

I think that NVIDIA uses different vent design on the IO shield,other than that it looks fairly accurate.

Good Job 

 

Example:

Yeah, I should have included at least one USB type C, or maybe 2, I can still change it, but now it too late to catch the hype train.

[Alan Mattano]

Stacking multiple RTX 3080 prototypes, the air will pass through. This system will allow the air passing through one GPU to other. This Is a special feature and also fan blades are surrounded by a turbo ring winglet.

My sensation is that  the system is a push from one side and pull from the other side because the camber airfoil profile of the blade looks reversed.

 

It looks to me that can be a pull extractor and not an intake. In any case air will pass through one GPU to other

If not the blade will spin in opposite direction.

 

 

[RejZoR]

@Alan Mattano

 

Something doesn't add up. I've rotated the top card to be oriented the same way as bottom one (the way you'd fit it into a case) and it seems to be inverted cooler placement on the top card. The lack of PCIe connector bothered me. Because it's on the other side and the cooler is hiding it!

 

Does this mean if you'll go with SLi you'll have to pick asymmetrically designed cards so they fit together better, so they don't suffocate each other's cooling? Just not sure which would go on top and which on the bottom. This is confusing.

[Alan Mattano]

@RejZoR Hi, I think what you mean is that you flip it (using a horizontal mirror).

If real, then this are tow different assembly layout prototypes proposals and maybe not intended to work together or work at all. This is a standard processes any industry. So they can have different hiding fans on the other side with different airflow. This are two different blades.

they are only protos.

It could be that the bottom GPU is blowing hot air to the memory (and CPU). And they whant to test that.

and the top GPU (on the original picture) the one fans is extracting air out from memory and cpu.

 

Lucubration presumption: On sketches could it be reversible. Or original designer intention was reversible and loos that ability in the feasibility production process.

We do not know.

[CarlBar]

Good catch @RejZoR. And it's not just an issue in that sense, it would mean the cooler shroud would be raised up from the PCI-E slot in the direction of the CPU cooler. Given clearance issues MB's can have somtimes with larger CPU coolers and the PCI-E slot thats not a good configuration.

 

Conversely if i'm wrong and the bottom card has the PCIE slot on the side closest to the camera that isn't an issue. However it's a bit odd IMO that the name is stamped then on what would effectively be the backplate. Usually it's on the frontplate and/or side most prominently.

 

Also hasn't NVIDIA effectively abandoned SLI?

[Bombastinator]

17 hours ago, RejZoR said:

@Alan Mattano

 

Something doesn't add up. I've rotated the top card to be oriented the same way as bottom one (the way you'd fit it into a case) and it seems to be inverted cooler placement on the top card. The lack of PCIe connector bothered me. Because it's on the other side and the cooler is hiding it!

 

Does this mean if you'll go with SLi you'll have to pick asymmetrically designed cards so they fit together better, so they don't suffocate each other's cooling? Just not sure which would go on top and which on the bottom. This is confusing.

This flipping about does display a problem.  With the chassis connector in the same orientation the pcie connector is now on opposite sides for each cooler.
 

 The two coolers can’t be the same.  More indication that there may be something not right about this leak.

 

This may just be a restatement if what you said.

[RejZoR]

5 hours ago, Alan Mattano said:

@RejZoR Hi, I think what you mean is that you flip it (using a horizontal mirror).

If real, then this are tow different assembly layout prototypes proposals and maybe not intended to work together or work at all. This is a standard processes any industry. So they can have different hiding fans on the other side with different airflow. This are two different blades.

they are only protos.

It could be that the bottom GPU is blowing hot air to the memory (and CPU). And they whant to test that.

and the top GPU (on the original picture) the one fans is extracting air out from memory and cpu.

 

Lucubration presumption: On sketches could it be reversible. Or original designer intention was reversible and loos that ability in the feasibility production process.

We do not know.

Just noticed different curvature on fan blades and I don't think it's just perspective at work. Also that ring around blades seems to suggest it has to be a directional fan and not one that would have to push air sideways through fins.

4 hours ago, CarlBar said:

Good catch @RejZoR. And it's not just an issue in that sense, it would mean the cooler shroud would be raised up from the PCI-E slot in the direction of the CPU cooler. Given clearance issues MB's can have somtimes with larger CPU coolers and the PCI-E slot thats not a good configuration.

 

Conversely if i'm wrong and the bottom card has the PCIE slot on the side closest to the camera that isn't an issue. However it's a bit odd IMO that the name is stamped then on what would effectively be the backplate. Usually it's on the frontplate and/or side most prominently.

 

Also hasn't NVIDIA effectively abandoned SLI?

NVIDIA abandoned multi-SLi. They still go with dual SLi. It kinda still makes sense as no one can make single GPU faster than two. Even if it doesn't scale perfectly.

[BiG StroOnZ]

NVIDIA GeForce RTX 3080 cooler could cost up to 150 USD alone:

Quote

 

https://videocardz.com/newz/nvidia-geforce-rtx-3080-cooler-could-cost-up-to-150-usd-alone

[porina]

The edited image by RejZor appears to be a flip (mirror, not rotate), so thanks for adding more confusion to this topic.

 

That render of how it could be constructed, personally I don't think they would need that fat a PCB as it only needs to carry power. There's a fair chance they might use wires instead, which would block the through airflow less.

[emosun]

i think it looks fine what were the complaints?

I'd be more pissed if it was just another stupid black rectangle like everything else in this world is

[ESPImperium]

The design is an okay looking one. Looks as if it will be a dust magnet. As long as there's some great 2 slot designs coming from the AIBs ill be fine. My machine is so damn tight that i can only have a 2S GPU, otherwise I'm loosing a USB card or a secondary PCIe M.2. As i only have a mATX machine.

[RejZoR]

13 hours ago, porina said:

The edited image by RejZor appears to be a flip (mirror, not rotate), so thanks for adding more confusion to this topic.

 

That render of how it could be constructed, personally I don't think they would need that fat a PCB as it only needs to carry power. There's a fair chance they might use wires instead, which would block the through airflow less.

What confusion? I just rotated it visually so it is positioned the same as the below one (which is in a position of installation into PCIe slot). See the mounting bracket for the rear of the case...

[porina]

1 minute ago, RejZoR said:

What confusion? I just rotated it visually so it is positioned the same as the below one (which is in a position of installation into PCIe slot). See the mounting bracket for the rear of the case...

It wasn't rotated, it was mirrored. So it doesn't represent what the original image shows. If you wanted a rotated one, it was already present as the lower GPU in the original image.

[RejZoR]

1 minute ago, porina said:

It wasn't rotated, it was mirrored. So it doesn't represent what the original image shows. If you wanted a rotated one, it was already present as the lower GPU in the original image.

I think you're misunderstanding description here. I've rotated a physical card. If image was mirrored or not is entirely irrelevant. Mirroring it doesn't magically change the alignment of the mounting bracket in relation to the backplate, PCB or the cooler. They are all still in the same fixed positions relative to the bracket. I just oriented bracket to the same position as the bottom card.  Which revealed that the cards are not oriented the same way (one has backplate facing up like all the graphic cards and the above one had backplate facing down (or cooler up however you want to call it).

[porina]

6 minutes ago, RejZoR said:

I think you're misunderstanding description here. I've rotated a physical card. If image was mirrored or not is entirely irrelevant.

Correct, I don't understand what you are trying to show. The original image, showing the two sides of presumably the same card, makes perfect logical sense. What you are doing, I'm lost.

[Bombastinator]

9 minutes ago, porina said:

Correct, I don't understand what you are trying to show. The original image, showing the two sides of presumably the same card, makes perfect logical sense. What you are doing, I'm lost.

I think the point that presuming it is the same card might not be right.   It doesn’t automatically mean it’s not.  It’s a fly in the ointment though.  We may not know as much as we think we do about how the thing is put together.

[Bombastinator]

2 minutes ago, Jack_of_all_Trades said:

 

I just dont see what the fan in the back is connected to cool.

 

Also the way this "interpretation" is made, but if you focus on the white shroud part where it says "geforce rtx" , it seems it physically separates the blue and red , so no air will be out or through the red heat fins ? It just does not make much sense or am I just that stupid, or maybe I am not supposed to think that critically of that image since its just an interpretation ?

There seem to be several sorts of suppositions about how these things are put together.  I personally view it as one theory.

[RejZoR]

@porina

Actually, you are correct. After inspecting the image again and checking positioning I made a mistake there because I took a shortcut and flipped the image using mirror instead of rotating it enough times. I've corrected the image and now they are actually positioned the same way, top one is just lifted so you see it from the "bottom" side. If you "lay" it down keeping the orientation it would be placed the same. PCB, PCIe connector and backplate (which isn't visible) are now in the correct position/side.

 

Still, when two cards are sandwiched in SLi, one fan will be suffocated (the left one). Right one will either have a pass through from above one or just dragging some air out of those fins above it.

 

Also I hope everyone understands I was rotating the top card so I get it positioned the same as the bottom one on the original image.