• 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.

NVIDIA 4000 Series

Associate
Joined
21 Apr 2007
Posts
2,076
The NVIDIA Ada Lovelace AD102 GPU will feature up to 12 GPC (Graphics Processing Clusters). This is an increase of 70% versus GA102 which features only 7 GPCs.
however GA102 was in the 3080 product sku and AD102 is thought not to be in the 4080

beware of the smoke and mirrors, remember Nvidia invented the 90 class sku during Ampere purely to segment the market
 
Soldato
Joined
31 Oct 2002
Posts
8,670
Permabanned
Joined
31 Aug 2013
Posts
3,364
Location
Scotland

kopite7kimi
@kopite7kimi
OK, let's do a new summary.
RTX 4090, AD102-300, 16128FP32, 21Gbps 24G GDDR6X, 450W, ~2x3090.
I am disappointed with RDNA3.
That's all.

What news did I miss on RDNA3?
 
Soldato
Joined
28 Oct 2009
Posts
2,738
Location
Birmingham
just getting sucked into Rumors again yeah good luck with that we never had anything thats been twice performance from gen to gen ? at most its been around 25% ??

wouldnt make sense for jump like that , better to do it by step by step , with AMD now competing maybe will get more than usual increase ;)
 
Last edited:
Soldato
Joined
6 Feb 2019
Posts
10,838

4090 = twice as fast as 3090 (rumour)

Take my money! :D

What happened to 600w?

Was MLID just making up stuff again for the last 6 months?
 
Soldato
Joined
6 Feb 2019
Posts
10,838
What news did I miss on RDNA3?

The only new rumour I've seen is that AMD is launching up the stack instead of done with a big launch window starting with Navi 33 and by the time 31 arrives it's 6 months after the RTX4090 has launched


If that's not what he's referring to the I have no idea but the crux of the latest rumour is that the 7900xt/7950xt won't be arriving until March 2023
 
Associate
Joined
30 Dec 2013
Posts
1,813
Location
Liverpool
I don't know where people are getting the 25% increase from. If we compare previous models based on the NAME only and not the gpu there's always been more than 25% increase.

980 > 780 more than 25%
980 Ti > 780 Ti more than 25%
1080 > 980 more than 25%
So on and so forth.

I'd expect the 4080 to be a good 60-70% faster than the 3080 which will put it at a good 30% faster than the 3090 Ti and if priced less than £700 should be able to do 4k144hz with a few tweaks here and there
 
Associate
Joined
16 Oct 2012
Posts
147
I've been sceptical of the rumours up until now but 450W for the 4090 is much more believable, that combined with 15% of the cuda cores being left on the table so that there is room for both a 3090TI and a Titan is very NVIDA.

One thing that not many people have mentioned is that while the rumoured FP32 performance is going to more or less double, the ram bandwith is only increasing by about 10%. On the surface you might think that could limit performance significantly but i have a feeling that it wont be all that bad. Looking at traditional Raster based games my guess is that it probably wont throttle performance at all, in God of War for example the memory bus utilization on a 3090 is only around 35%, so that leaves plenty of scope for more performance. On the other hand in ray traced games regularly see well above 50% bus utilisation, so i wouldnt be completely surprised if we see some strange scaling vs Ampere in RT games. That being said NVIDIA have significantly increased the on die cache for Lovelace similar to infinity cache for AMD on the 6900 series which make do just fine with a 256bit bus and 18GB/s memory. It's not unreasonable to think similar is possible for NVIDIA.
 
Soldato
Joined
31 Oct 2002
Posts
8,670
I don't know where people are getting the 25% increase from. If we compare previous models based on the NAME only and not the gpu there's always been more than 25% increase.

980 > 780 more than 25%
980 Ti > 780 Ti more than 25%
1080 > 980 more than 25%
So on and so forth.

I'd expect the 4080 to be a good 60-70% faster than the 3080 which will put it at a good 30% faster than the 3090 Ti and if priced less than £700 should be able to do 4k144hz with a few tweaks here and there

You think the 4080 will be less than £700? Be very lucky for it to be under £1000 IMO.

3080 at £650 will never happen again, given the inflationary climate we're in, which is exacerbated by global supply issues etc etc.
 
Soldato
Joined
28 Oct 2009
Posts
2,738
Location
Birmingham
I don't know where people are getting the 25% increase from. If we compare previous models based on the NAME only and not the gpu there's always been more than 25% increase.

980 > 780 more than 25%
980 Ti > 780 Ti more than 25%
1080 > 980 more than 25%
So on and so forth.

I'd expect the 4080 to be a good 60-70% faster than the 3080 which will put it at a good 30% faster than the 3090 Ti and if priced less than £700 should be able to do 4k144hz with a few tweaks here and there
one of Jay2cents rececnt uploads was saying typically been 25-30% generation over generation on the top tiers , this in actual games also they artificially do it and not want to push to far too soon , but with AMD getting closer might start pushing more

anything below 50% I will be disappointed
 
Last edited:
Soldato
Joined
16 Jun 2004
Posts
2,976
You think the 4080 will be less than £700? Be very lucky for it to be under £1000 IMO.

I'm pretty sure we'll see a £749-£799 sticker price for the 4080 FE models, which will leave the AIBs plenty of room to price their custom cards upto a £1000 or so.
 
Soldato
Joined
21 Jul 2005
Posts
16,751
Location
N.Ireland
That really depends on how much of a % uplift we see from Ada, followed by how much this expensive vram seems to be. If they are not being greedy, you could see the FE pencil in at £699 although I would second @Cooper prices if the modules cant be stingy again.
 
Caporegime
Joined
18 Oct 2002
Posts
26,483
Location
y0 Mommas a**
They switched Ada to Linda at the last second...

image.png
 
Top Bottom