• 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

Man of Honour
Joined
13 Oct 2006
Posts
91,171
There isn't really a standard length of time - depends on supply and how much or not it might be a "paper" launch. Generally from when the CEO of the company stands up on stage waving the card around you are looking at 1-2 months before they appear in pre-builds but supply might be another matter.
 

HRL

HRL

Soldato
Joined
22 Nov 2005
Posts
3,028
Location
Devon
Errrm probably 90% of the buying public? You're making the mistake of believing that everyone is an enthusiast like us on here, whereas the truth is that Joe Public (a massive part of the sales) has no idea when the new cards launch or just go out and buy what they need right now. We greatly overestimate this all the time because we're inside the bubble...

Fair point and you’re probably right. Still think it’s starting to smell like new cards are imminent though, at least from Nvidia.

Can’t see AIB’s knocking a $1K off their top tier cards just to be nice, like EVGA have just done, albeit I don’t doubt that their sales have dropped for a whole multitude of reasons.
 
Soldato
Joined
30 Aug 2014
Posts
5,966
35W seems very close. Does that mean the 4080 won't be as fast as we think or that the 4070 will be faster than expected?
I imagine Nvidia will find a way to segment them :cry:. There are plenty of things they can do to make a card slower and anyway the 4080 is rumoured to be based on the bigger 103 chip vs the 104 in the 4070 (the latter may be clocked higher to meet performance targets, hence the higher power draw as you go beyond the sweetspot for efficiency).
 
Soldato
Joined
15 Oct 2019
Posts
11,701
Location
Uk
Soldato
Joined
15 Oct 2019
Posts
11,701
Location
Uk
Information changes like the wind right now. Last one I watched they inferred the new lineup had weaker bandwidth this time although it could be offset using better ipc, bigger cache etc. Only going to see once people test them.
The extra cache worked for AMD last time which allowed them to use a cut down bus with more VRAM so it's not supprising if nvidia choose to go down that route as the flagship will need more VRAM this time around.
 
Soldato
Joined
15 Oct 2019
Posts
11,701
Location
Uk
What makes you say that?
There will be more games that push past 10gb over the next couple of years and while the 3080 will be less effected by this as it will need to lower settings anyway to stay above 60fps which in turn will reduce VRAM consumption the 4080 should stay above 60fps without the need to drop settings so would need atleast 12gb.
 
Man of Honour
Joined
22 Jun 2006
Posts
11,675
There will be more games that push past 10gb over the next couple of years and while the 3080 will be less effected by this as it will need to lower settings anyway to stay above 60fps which in turn will reduce VRAM consumption the 4080 should stay above 60fps without the need to drop settings so would need atleast 12gb.

Seeing as AMD chose cache this is probably a stupid question, but what costs more, cache, or vram? I know that shrinks can offer more die space, but presumably using big chunks for cache has implications for performance which vram does not. In the Intel Arc video GN had awhile back, I think the engineer said using a bigger bus makes the die bigger so it makes everything cost more.
 
Soldato
Joined
27 Feb 2015
Posts
12,621
Caporegime
Joined
22 Nov 2005
Posts
45,288
Is there actually any leakers or just sites fighting for eyeballs making stuff up?
The age of clickbait, any myths, rumours or BS can fly and make the same $$$ as if you did a proper good video.


everything is just going along with whatever rumour or creating their own for clicks and all these "sources" are figments of their imagination.
how can soo many sources be soo wrong in soo little time...


we've gone from like 800watts > 600 > 400 > 300 in the space of like a week


Gamers Nexus seems to be the only one not putting out any old crap rumours for content so far?
 
Associate
Joined
22 Mar 2015
Posts
413
Location
Kent, UK; Brooklyn, NYC, NY
The extra cache worked for AMD last time which allowed them to use a cut down bus with more VRAM so it's not supprising if nvidia choose to go down that route as the flagship will need more VRAM this time around.
They're going to have to rely on cache to mitigate the 256-bus. Memory speed should make up for the rest. And I'm gonna venture to guess nV will be much less vocal about 8k gaming this gen.
 
Soldato
Joined
12 May 2014
Posts
5,237
Seeing as AMD chose cache this is probably a stupid question, but what costs more, cache, or vram? I know that shrinks can offer more die space, but presumably using big chunks for cache has implications for performance which vram does not. In the Intel Arc video GN had awhile back, I think the engineer said using a bigger bus makes the die bigger so it makes everything cost more.
A bigger bus also uses more power.
 
Back
Top Bottom