• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Geforce GTX1180/2080 Speculation thread

Gerard said:
Feel sorry for that card after he tore it apart :( Regardless of price it looks good and seems to be very solidly put together.
Click to expand...

Indeed he should be shot for destroying it. :mad:

Mate that cooler is beefier than the one used on the Vega 64 Nitro+ and the GTX1080Ti Xtreme. Cards designed to deal with 330W power consumption (both).
And that make me thing that it will run very power hungry
 
Wow, if true the two different SKUs thing is interesting. Could be all sorts, better chips for some AIBs / higher tier cards. Chips for NV only. For reviews. Maybe for an as yet undisclosed ‘Titan’ 2085 jobby.

Doesn’t necessarily spell doom, it could even be that the A’s are for us gamers and that the non are for some content creation card yet to be released.

It’s dissapointing though that it’s another thing NV aren’t telling us though (if true), I would have thought that’s something you bring up at the launch presentation. This whole release has been super weird imo.

EDIT : Just watched the tear down. Not only is that an absolutely stunning cooler, but it looks very well put together. Their best yet by a large margin.
 
Last edited:
Panos said:
And that make me thing that it will run very power hungry
Click to expand...

Which is pretty much a given with the size of those cores the question is does it have the performance to match and can it adaptively manage power when not using the Tensor/RTX functionality that are part of why the cores are so big.
 
Panos said:
Indeed he should be shot for destroying it. :mad:

Mate that cooler is beefier than the one used on the Vega 64 Nitro+ and the GTX1080Ti Xtreme. Cards designed to deal with 330W power consumption (both).
And that make me thing that it will run very power hungry
Click to expand...

It's likely to be around 80% faster than the Vega 64, so even if it hits 330 W, that's pretty reasonable in context. :)
 
JediFragger said:
775mm sq, that sucker is gonna be HOT when it's all firing!! :eek:
Click to expand...

They seemed pretty confident on presentation day. I’m reckoning there’s a chance the FE cooler does a very good job. What with the wattage figures and die size I’m probably wrong.

EDIT: and so many 2.5 - 3 slot coolers. You’re right mate, I’ll be quiet.
 
Panos said:
Indeed he should be shot for destroying it. :mad:

Mate that cooler is beefier than the one used on the Vega 64 Nitro+ and the GTX1080Ti Xtreme. Cards designed to deal with 330W power consumption (both).
And that make me thing that it will run very power hungry
Click to expand...
Why they gonna slam Aio on it like instant like always...
Air cooling ?? Who cares :D Not had AC in gpu since eee cant remember had water on my ATI 4850 so how many eyars is that.

Besides he could always do Smash my RTX.com like they used to do :)
 
Last edited:
Panos said:
TSMC 7nm doesn't face delays though. :)
On the contrary, the process is so good, that AMD is releasing earlier than planned (by 6 months) some of their products. Like the Instinct Vega 20 cards. While EPYC Rome is succesfully been sampled for 4 months now.
Click to expand...

TSMC 7nm will face delays if the yields are at 30-40% just like it had happened with both Samsung and TSMC's 10nm back in 2017 saw poor yield 30-40% that delayed Apple A11X chips shipments for 2017 Ipad models. Samsung and TSMC 10nm biggest customer is Apple.

https://techspecs.blog/blog/2017/4/5/why-the-new-ipads-are-delayed

AMD had been planned Vega 20 cards, EPYC Rome and Zen 2 3000 series APUs/CPUs to use both TSMC and GlobalFoundries's 7nm process to increase supply and demand but as you probably not aware Globalfoundries cancelled 7nm process last month mean AMD will have to use TSMC as only 1 source could see long wait for shipments or ship fewer chips.

Apple is TSMC biggest customer again on 7nm with A12 chip used 75% of 7nm capacity.

https://www.eetimes.com/document.asp?doc_id=1333615

I don't know what will happen in 2019 as it seem every company go to TSMC 7nm, Intel could be one of customer outsourced future chipset or low end CPU production to TSMC 7nm when both Intel 10nm and 14nm faced delays, shortage, poor yield or full capacity.
 
Panos said:
Yeah. That is going to be funny. Especially when people are pre-ordering but they have no idea if it will be Normal SKU or OC SKU.

Which reminds, Nvidia applying Intel tactics here with the "K series" but without letting clearly the users knowing what they buy until GPUZ is used....
Click to expand...

JediFragger said:
775mm sq, that sucker is gonna be HOT when it's all firing!! :eek:
Click to expand...

Said at the very beginning there was going to be a lot of flipping on heat output.;)

Can't believe he destroyed that cooler, but more disappointed he's on the nda list, now we don't know what's what with the differing chips because they are obviously not allowed to report it as it's under nda(and some said the nda was normal practice coz they signed one at work:o), tbph, whole PC gfx scene is leaving a severely sour taste...
 
Have the Final Fantasy RTX 2080/2080ti scores been shared here yet?

in case they havent here they are in spoiler.
http://benchmark.finalfantasyxv.com/result/

pkSpE5O.png

PBSAoT3.png

C4PbqKc.jpg
 
Panos said:
Indeed he should be shot for destroying it. :mad:

Mate that cooler is beefier than the one used on the Vega 64 Nitro+ and the GTX1080Ti Xtreme. Cards designed to deal with 330W power consumption (both).
And that make me thing that it will run very power hungry
Click to expand...


Be interesting to see FE temps vs the various other cards, 2 slot vs almost 3 on some cards. And vapour chamber vs what seems to be basically heatpipes and regular aluminium fins on others.
 
An explanation of dlss from hardware unboxed says there are two modes.

The first mode works by hitting the resolution you want by upscaling a lower resolution image. This is likely the mode that gives performance uplifts.

Then there is the 2x mode which is basically intelligent super sampling. So renders at the resolution you want but uses information from a super sampled image created using the tensor cores (ground truth is a 64x super sampled image).
 
Panos said:
I believe you do have a 1000W platinum PSU yes? So take a deep breath plug the third connector to it, and don't turn on the central heating this winter.
You should be OK :p

Another good service by Nvidia. You warm your house while playing games during winter. No need for boiler and gas bills :p
Click to expand...

Yet another thing they have beaten AMD at then, they heat your house better now as well. ;)
 
WantoN said:
Wow, if true the two different SKUs thing is interesting. Could be all sorts, better chips for some AIBs / higher tier cards. Chips for NV only. For reviews. Maybe for an as yet undisclosed ‘Titan’ 2085 jobby.

Doesn’t necessarily spell doom, it could even be that the A’s are for us gamers and that the non are for some content creation card yet to be released.

It’s dissapointing though that it’s another thing NV aren’t telling us though (if true), I would have thought that’s something you bring up at the launch presentation. This whole release has been super weird imo.
Click to expand...
From earlier in the thread, it appears to be quite simply any overclocked card has an "A" chip (incL FE). Reference cards are "standard" chips.
 
You must log in or register to reply here.
Back
Top Bottom