-
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.
Blackwell gpus
- Thread starter scubes
- Start date
Associate
looks like a well rounded architecture -the rtx 5090, just hoping for jensen to not cut this any further
nvidia might bring back a complex scheduler with how the 4090 was being severely limited by the CPU in competitive gaming scenarios
so perhaps the SM organization might change a bit
overall, looks solid - just waiting for jensen to start his keynote with "my ampere gamer friends its safe.. "
nvidia might bring back a complex scheduler with how the 4090 was being severely limited by the CPU in competitive gaming scenarios
so perhaps the SM organization might change a bit
overall, looks solid - just waiting for jensen to start his keynote with "my ampere gamer friends its safe.. "
That 5080 with only 16GB will be another sales flop priced anything over £799, IMO.
The marketplace is fundamentally different to what it was in late 2022 and that pricing structure was an aberration, that won't fly again.
Despite what everyone is fearing I expect the 5090 price to remain pretty much static and Nvidia will pull another '3080 moment' with the 5080.
The marketplace is fundamentally different to what it was in late 2022 and that pricing structure was an aberration, that won't fly again.
Despite what everyone is fearing I expect the 5090 price to remain pretty much static and Nvidia will pull another '3080 moment' with the 5080.
Caporegime
You're underestimating the FOMOcels, who will be happy to hand over their hard-earned cash once again, because one simply can't enjoy a game without a 5090 and pixel creep.
As long as the halo card is selling well, they'll be happy to take the hit on the 5080. If we adjust the 3080 RRP for inflation, it's about $850 today and I really can't see nVidia doing that.
Last edited:
I don't really pay any attention to the 90 class and its effect on the GPU market below it.
Forums like this amplify their impact way beyond their actual effect, when they're really only a tiny percent of the GPU market, that only appeal to the very few.
Associate
- Joined
- 13 Sep 2010
- Posts
- 2,177
Yes, because it uses 2GB ICs...
"With 3GB ICs, the RTX 5080 would be able to offer 24GB of VRAM without adding or reducing the memory chip count, enabling it to retain its 256-bit wide interface. Depending on how Nvidia wants to operate in the age of GDDR7, it could also play the same trick with the RTX 5090 and give it 48GB of memory with 3GB ICs."
source
"But on top of that, 24Gb GDDR7 chips are also going into production, for which the effective capacity would be 3GB per chip. And that will allow for graphics cards with 50% more memory capacity. So in addition to the previously cited numbers, it will also be possible to have 12GB graphics cards with a 128-bit bus, 18GB graphics cards with a 192-bit bus, 24GB graphics cards with a 256-bit bus, or a capacity of 36GB for GPUs with a 384-bit bus width. A clamshell design would double those again."
source
Edit: Sorry, I thought in your original post you were saying I was wrong about the 36/48 capacities for 3GB chips, which you weren't!
Last edited:
Caporegime
You may not be paying attention, but what we saw last time had a huge impact on the GPU market below it, particularly the card down from it, the 4080.
Everything below it was pulled up, and they priced the 4080 within reach of a 4090, triggering a "the more you buy, the more you save" kind of mentality. Although we can only speculate at this stage, it looks like the performance gap between the 5090 and the 5080 is going to be huge again.
I'll eat my hat if nVidia release the 5080 below $1k.
Last edited:
Associate
- Joined
- 2 Feb 2009
- Posts
- 750
Associate
QC is another big concern around new GDDR7 ram modules - which means the cards arent coming out any sooner. but hope to see some activity around november end, jan is just too late
I doubt they'll have 3gb chips ready at launch though with is probably why the 5090 is being rumoured with a 512 bus and 32gb and not a 384 bus and 36gb, it'll need 16 VRAM chips which will be difficult to fit all on one side unless the chips are smaller.
It doesn't but it makes it more difficult to cool when they're on the rear as was the case with the 3090.
Associate
- Joined
- 13 Sep 2010
- Posts
- 2,177
Yep, absolutely not expecting 3GB chips at launch, possibly later but more likely the next gen or 2 after (they'll be wanting those high cap chips for datacentre/workstation cards).
It will have to be 16 chips though won't it? I assume they'll either do 4 around each side of the GPU (bit weird how 3090/4090 are a 4+4+3+1 layout, I assume there are reasons for that lol).
3GB memory chips aren't going to be physically bigger than 2GB ones as far as I know (thought the increase would be down to lithography process)
It will have to be 16 chips though won't it? I assume they'll either do 4 around each side of the GPU (bit weird how 3090/4090 are a 4+4+3+1 layout, I assume there are reasons for that lol).
3GB memory chips aren't going to be physically bigger than 2GB ones as far as I know (thought the increase would be down to lithography process)
