NVIDIA ‘Ampere’ 8nm Graphics Cards

[Razor Time]

NP .

Not so much worth it for me anymore. A lot of hassle to fit for next to nothing performance wise. And a big pain to strip down and test in another oc should an issue arise. My pc is a distance away now so again noise isn’t much of an issue. When it was close the system was quiet bar the coil whine for me which I think blocks make worse so wanting to stick with an air cooled card now anyway.

Thatys fair enough, it is a huge amount of hassle to add to a system. I am a bit OCD when it comes to noise so always go through the pain of blocking a card.

 

[dreigns]

It's worth mentioning that for most of the hardware monitors they're only reporting what the game has reserved in vRAM, they are not listing real vRAM usage. For most games the game engine allocates a certain amount space in vRAM (typically an over estimate of what it needs) and then internally manages what that space is used for. Other than dev tools baked into game engines which give insight into internal vRAM usage the only other tool I'm aware of that will report on actual vRAM usage is Special K, it was recently removed from steam due to some drama but you can find downloads and discussion here https://discourse.differentk.fyi/

I'd be interested in what you get if you run Special K and compare memory usage. My bet is it's below 10Gb and actually more than likely close to or below 8Gb.

Lastly I'd be interested in actual frame rates at the settings you posted above that exceed 10Gb, the averages and minimums if possible.

Yup I was gonna point that out just having gpu-z open without anything running is showing a use of 2.1-2.5gig.

Also I looked at AC:O and maxed out at 4k with everything at max it says it will use 5266gig not bad for a game that size.

On another note SOTTR running at 4k Maxed no RTX on myn is getting 41 Fps in DX12 mode so a 3080 at 30-40% should be getting 55-60fps as I think they have done some magic with DX12, I also tested with DX11 and had that running at 58fps.
These are the average fps taken.

For reference I have a AMD 3900X ( Not Overclocked ) and DDR4 3200 32gig so not the fastest cpu for games.

 

[LoadsaMoney]

Same size as that god awful 5700, thats been dropped on its edge then

 

[tyler_jrb]

Thatys fair enough, it is a huge amount of hassle to add to a system. I am a bit OCD when it comes to noise so always go through the pain of blocking a card.

Yeah I was very much the same I’ve had several 1080Tis and 2080Tis because of it. One day when I’d had enough, literally about to sell the pc I just emptied my wall cupboard as a test and put it in there. Obviously not as nice as being on the desk but fans max 1000rpm door cracked about 1inch. No noise at all as it’s about 10ft away. Before the whine was sending me crazy. Best thing I’ve done.

A good thing as it means I can put anything in there and not have to worry about the noise it makes.

 

[Kaapstad]

Yup I was gonna point that out just having gpu-z open without anything running is showing a use of 2.1-2.5gig.

Also I looked at AC:O and maxed out at 4k with everything at max it says it will use 5266gig not bad for a game that size.

On another note SOTTR running at 4k Maxed no RTX on myn is getting 41 Fps in DX12 mode so a 3080 at 30-40% should be getting 55-60fps as I think they have done some magic with DX12, I also tested with DX11 and had that running at 58fps.
These are the average fps taken.

For reference I have a AMD 3900X ( Not Overclocked ) and DDR4 3200 32gig so not the fastest cpu for games.

You need to rerun SOTTR using the settings I posted earlier in the thread.

If you don't use RTX then yes you will get a higher fps but that defeats the point of buying a 3000 series card.

 

[Bindlestiff]

Take my money I have had a reshuffle, new case and AOI to the front due to the exhaust fan.
f5 f5 f5

You have one?

 

[TNA]

You have one?

Not sure how you got that from that post

 

[Razor Time]

Yeah I was very much the same I’ve had several 1080Tis and 2080Tis because of it. One day when I’d had enough, literally about to sell the pc I just emptied my wall cupboard as a test and put it in there. Obviously not as nice as being on the desk but fans max 1000rpm door cracked about 1inch. No noise at all as it’s about 10ft away. Before the whine was sending me crazy. Best thing I’ve done.

A good thing as it means I can put anything in there and not have to worry about the noise it makes.

Its funny the lenghts some of go to due to pc noise, whilst others dont care at all, glad you found an easy solution. Althought I have to say I do remember some of your 900D builds and they looked very good.

 

[jonneymendoza]

@jonneymendoza just saw your updated sig, 128gb RAM? Damn boy, whatcha need that much ram fo?

LOL.

I do some photo and video editing as well as coding.

But the real reason was that i accidently bought two sticks of 32gb instead of 4 16gb sticks(found out my socket alows quad channel) and i could have lived with just 64gb lol

 

[benbird7]

You have one?

No see that image posted, I would like one in a weeks time. But fear the demand is going to far outweigh the supply.
Hoping to see some decent reviews before release with likely increases for VR, currently got the 1080ti for flying around in my rift s

 

[Bravetart]

Its funny the lenghts some of go to due to pc noise, whilst others dont care at all, glad you found an easy solution. Althought I have to say I do remember some of your 900D builds and they looked very good.

I'm all for a quiet PC but generally have a headset on, which negates the need to a degree... The fans in my H100i v2 become annoyingly loud at anything 1rpm over 1000rpm though...fortunately easily solved by setting a custom curve, but annoying all the same.

 

[TNA]

LOL.

I do some photo and video editing as well as coding.

But the real reason was that i accidently bought two sticks of 32gb instead of 4 16gb sticks(found out my socket alows quad channel) and i could have lived with just 64gb lol

 

[4k4uNuNu]

A week until launch and no benchmarks or reviews, no pre-orders, nerfed vram on the 3080, no decent upgrade path for 2080ti owners and a massive value loss on that card. They also seem power hungry which is a concern for 650w users and overclockers.

I won't lie, it's leaving a bitter taste giving over a grand to nvidia two years ago. My next graphics card might be an AMD, just out of spite lol

 

[Kelt]

That is actually very unappealing seeing it in a case.

I mentioned it an earlier post, but I dislike the look of that power cable.

What were Nvidia thinking, placing it there.

The rest of the card looks fine, but that cable placement ruins the overall look.

 

[TNA]

A week until launch and no benchmarks or reviews, no pre-orders, nerfed vram on the 3080, no decent upgrade path for 2080ti owners and a massive value loss on that card. They also seem power hungry which is a concern for 650w users and overclockers.

I won't lie, it's leaving a bitter taste giving over a grand to nvidia two years ago. My next graphics card might be an AMD, just out of spite lol

Do what others do, reward them and buy a 3090

 

[varkanoid]

Yeah but then the fan blade isn't cutting into the air and driving it up?

the sample we saw ages ago looked like it had the fans on a peculiar way are we saying we were right the fans dont make sense airflow wise ?

 

[ALXAndy]

Its not though.

Lets say your baseline score is 100fps with the 2080 and the 3080 is 80% faster so is 180fps.

And lets go with the 2080ti being 35% faster than the 2080 so would be 135fps.

180-135= 45 fps

45/135 = 33%

So on your figures even when the 3080 is 80% faster than the 2080, its only 33% faster than the 2080ti.

Even in perfect RT?DLSS games where the 3080 is 100% faster than the 2080, the maths work out as follows:

200-135= 65fps

65/135= 48% faster.

So we will probably be looking at 33% faster than a 2080ti and maybe in ideal perfect settings, it might be up to 48% faster.

And that assumes they both clock the same (which isn't looking likely since EVGA has only gone with a 5% overclock on its FTW3 Ultra card which is very very conservative)

So yeah i can easily see the 3080 only being 20-30% faster than a 2080ti in general and up to 40% more in ideal conditions.

Well calculated. You missed one, though

30% more power than a 2080Ti.

2080Ti = 250w
3080 = 320w.

So 30% higher power consumption also.

 

[TNA]

I mentioned it an earlier post, but I dislike the look of that power cable.

What were Nvidia thinking, placing it there.

The rest of the card looks fine, but that cable placement ruins the overall look.

If I get it I will later on buy a single nice braided cable going from psu directly into it. I will do it from the side and I think that will look fine then.

 

[TonyTurbo78]

I mentioned it an earlier post, but I dislike the look of that power cable.

What were Nvidia thinking, placing it there.

The rest of the card looks fine, but that cable placement ruins the overall look.

Indeed. Its such a stupid position for it, it makes no sense plonking it there. It don't matter which way you put it, having 2 cables or 1 splitter cable coming from that part of the card, it's directly in the middle of your view into the case. Plus, it'll cover up the LED logo on the card surely. Then think how crap it'll look with a sleek water block with that power connector there with two fat wires coming from it in direct view.

 

[Razor Time]

I'm all for a quiet PC but generally have a headset on, which negates the need to a degree... The fans in my H100i v2 become annoyingly loud at anything 1rpm over 1000rpm though...fortunately easily solved by setting a custom curve, but annoying all the same.

I have been through so many different fans, the best ones I found were the be quiet! Silent Wings 3. I'm not sure if you have tried them but worth a look if you like low noise and decent performance.