• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Video Memory Sizes Set to Swell as NVIDIA Readies 20GB and 24GB GeForce Amperes

btarunr

Editor & Senior Moderator
Staff member
Joined
Oct 9, 2007
Messages
47,670 (7.43/day)
Location
Dublin, Ireland
System Name RBMK-1000
Processor AMD Ryzen 7 5700G
Motherboard Gigabyte B550 AORUS Elite V2
Cooling DeepCool Gammax L240 V2
Memory 2x 16GB DDR4-3200
Video Card(s) Galax RTX 4070 Ti EX
Storage Samsung 990 1TB
Display(s) BenQ 1440p 60 Hz 27-inch
Case Corsair Carbide 100R
Audio Device(s) ASUS SupremeFX S1220A
Power Supply Cooler Master MWE Gold 650W
Mouse ASUS ROG Strix Impact
Keyboard Gamdias Hermes E2
Software Windows 11 Pro
NVIDIA's GeForce RTX 20-series "Turing" graphics card series did not increase video memory sizes in comparison to GeForce GTX 10-series "Pascal," although the memory itself is faster on account of GDDR6. This could change with the GeForce RTX 30-series "Ampere," as the company looks to increase memory sizes across the board in a bid to shore up ray-tracing performance. WCCFTech has learned that in addition to a variety of strange new memory bus widths, such as 320-bit, NVIDIA could introduce certain higher variants of its RTX 30-series cards with video memory sizes as high as 20 GB and 24 GB.

Memory sizes of 20 GB or 24 GB aren't new for NVIDIA's professional-segment Quadro products, but it's certainly new for GeForce, with only the company's TITAN-series products breaking the 20 GB-mark at prices due north of $2,000. Much of NVIDIA's high-end appears to be resting on segmentation of the PG132 common board design, coupled with the GA102 silicon, from which the company could carve out several SKUs spaced far apart in the company's product stack. NVIDIA's next-generation GeForce "Ampere" family is expected to debut in September 2020, with product launches in the higher-end running through late-Q3 and Q4 of 2020.



View at TechPowerUp Main Site
 
NVIDIA's G80 GPU also featured 320bit bus... so it is not anything new, check your facts.
 
This is just regurgitating old rumors isn't it? Is TPU really that keen on becoming WCCFTech? Its not exactly a good thing, you know.
 
WCCF is the crappiest tech tabloid on the net. They just keep pressing F5 on reddit and posting absolutely every rumor they find.

We're supposed to be jumping on the train right now and speculate along with the masses, will it be 12? Will it be 24? 16? ERmergerd what if its 13.

How about we don't, for once, and show TPU this is not the kind of headlines we need. @btarunr take note

I love how the PG board numbers are added (totally irrelevant) as if that somehow lends more credibility to it all.
 
Contradictory information usually means people are just guessing, when they are based on real information they tend to line up.
Until we have an actual leak of something, we shouldn't assume anything.
And for those wondering; Nvidia can certainly double memory if they want to, but it's getting late in the game to do such changes, as they can't be done after the cards are made.
 
I believe they manufacture some of their own :pimp:
They certainly do. WCCFTec is only useful when they refer to other credible sources.
Such websites and most tech news YouTube channels have the same business model; search Reddit, forums and bulletin boards to see what is trending, and generate something to feed the demand, with some occasional spicy or controversial twist.

Didn't ATI use 512bit bus? I believe it was on the HD2900XT.
Yes, so did GTX 280/285.
But I believe it's getting more and more challenging with higher memory speeds.
 
Yes, so did GTX 280/285.
But I believe it's getting more and more challenging with higher memory speeds.

Not challenging - unnecessary and expensive. Unnecessary because higher-speed memories and improved texture compression algorithms provide for more than enough bandwidth, expensive because more memory traces increase costs.
 
WCCF is the crappiest tech tabloid on the net. They just keep pressing F5 on reddit and posting absolutely every rumor they find.
going to wccftech for the articles is like buying playboy for the blurb
 
Not challenging - unnecessary and expensive. Unnecessary because higher-speed memories and improved texture compression algorithms provide for more than enough bandwidth, expensive because more memory traces increase costs.
By challenging, I mean challenging to fit that many chips on the PCB with traces of the same length. With GDDR6 they can no longer be scattered across the PCB.

Memory compression certainly saves some bandwidth and capacity, but this is mostly for sparse data. This is not to be confused with texture compression, which is applied in advance by the game.
 
This is just regurgitating old rumors isn't it? Is TPU really that keen on becoming WCCFTech? Its not exactly a good thing, you know.
More reputable sources like MLID have speculated on two 3080 variants. One with 20GB. Tho if Nvidia really does end up giving 3080Ti 24GB it will be a massive jump from current 11GB.
Personally i doubt it. I'll believe it when i see it.
 
NVIDIA's G80 GPU also featured 320bit bus... so it is not anything new, check your facts.

I feel this comes down to how you read it tbh.
Its like saying this graphics card is not new because we had graphics cards before.

Yes a 320bit bus has been done before, but not with this type of memory which calculates differently and its new information for these new cards.
 
going to wccftech for the articles is like buying playboy for the blurb
Wait. Playboys have articles in them? Damn, I didn't know that. I thought everyone went for the centerfold. :laugh:
 
We're supposed to be jumping on the train right now and speculate along with the masses, will it be 12? Will it be 24? 16? ERmergerd what if its 13.

How about we don't, for once, and show TPU this is not the kind of headlines we need. @btarunr take note

I love how the PG board numbers are added (totally irrelevant) as if that somehow lends more credibility to it all.

I've been beating this particular drum of quality over quantity for a while now. Good to know I'm not alone.

Wait. Playboys have articles in them? Damn, I didn't know that. I thought everyone went for the centerfold. :laugh:

Playboy had some amazing writers and articles back in the day: https://www.bbc.co.uk/news/entertainment-arts-41426419
 
Last edited:
Could see prices swelling as well then.
 
10 years ago GTX 460 had 1 and 2GB versions, and it made sense not only for SLI, since 1Gb proved to be starving the single cards,. Now we have 10 and 20GB, and those are the Nvlink enabled cards, in another 10 years, 100 and 200GB. The problem is that 30 series would be as hot as FERMI. Just wasted, better wait for 40 series, true 7nm or 5nm EUV.
 
Its strange how so far most Ampere topics seem to go in that direction. Very strange indeed.
The price. Pervert. :p :roll: :D :pimp:

What's really funny I didn't read and see the playboy discussion later in the thread until after I posted.

I really don't see a point in more than 12GB/16GB for this generation.
 
The price. Pervert. :p :roll: :D :pimp:

What's really funny I didn't read and see the playboy discussion later in the thread until after I posted.

I really don't see a point in more than 12GB/16GB for this generation.

More memory means they have an excuse for the insane prices.
 
Back
Top