Wednesday, November 4th 2020
NVIDIA Reportedly Working on GeForce RTX 3080 Ti Graphics Card with 20 GB GDDR6X VRAM
A leak from renowned (and usually on-point) leaker Kopite7kimi claims that NVIDIA has finally settled on new graphics cards to combat AMD's RX 6800 threat after all. After the company has been reported (and never confirmed) to be working on double-memory configurations for their RTX 3070 and RTX 3080 graphics cards (with 16 GB GDDR6 and 20 GB GDDR6X, respectively), the company is now reported to have settled for a 20 GB RTX 3080 Ti to face a (apparently; pending independent reviews) resurgent AMD.
The RTX 3080 Ti specs paint a card with the same CUDA core count as the RTX 3090, with 10496 FP32 cores over the same 320-bit memory bus as the RTX 3080. Kopite includes board and SKU numbers (PG133 SKU 15) along a new GPU codename: GA102-250. The performance differentiator against the RTX 3090 stands to be the memory amount, bus, and eventually core clockspeed; memory speed and board TGP are reported to mirror those of the RTX 3080, so some reduced clocks compared to that graphics card are expected. That amount of CUDA cores means NVIDIA is essentially divvying-up the same GA-102 die between its RTX 3090 (good luck finding one in stock) and the reported RTX 3080 Ti (so good luck finding one of those in stock as well, should the time come). It is unclear how pricing would work out for this SKU, but pricing comparable to that of the RX 6900 XT is the more sensible speculation. Take this report with the usual amount of NaCl.
Sources:
Kopite7kimi @ Twitter, via Videocardz
The RTX 3080 Ti specs paint a card with the same CUDA core count as the RTX 3090, with 10496 FP32 cores over the same 320-bit memory bus as the RTX 3080. Kopite includes board and SKU numbers (PG133 SKU 15) along a new GPU codename: GA102-250. The performance differentiator against the RTX 3090 stands to be the memory amount, bus, and eventually core clockspeed; memory speed and board TGP are reported to mirror those of the RTX 3080, so some reduced clocks compared to that graphics card are expected. That amount of CUDA cores means NVIDIA is essentially divvying-up the same GA-102 die between its RTX 3090 (good luck finding one in stock) and the reported RTX 3080 Ti (so good luck finding one of those in stock as well, should the time come). It is unclear how pricing would work out for this SKU, but pricing comparable to that of the RX 6900 XT is the more sensible speculation. Take this report with the usual amount of NaCl.
140 Comments on NVIDIA Reportedly Working on GeForce RTX 3080 Ti Graphics Card with 20 GB GDDR6X VRAM
Nvidia RTX QUADRO GEFORCE 3100X Ti NITRO SUPER TOURNAMENT ULTRA TITAN EDITION!
Clearly a knee-jerk reaction to AMD's line-up.
It is the other way around, Huang new Navi is coming so no more fun with pricing.
Compare it to turing, 2080Ti + 20-30% for $699, why so cheap?
And they will if history is any indication of things.
Lather, rinse, repeat.
All that will come with the £3000 Titan variant.
RIP Nvidia right? How could they possibly survive this flogging.
Even if Nvidia release a 3080ti now, the damage has already been done. Their availability and Nvidia's bad choices have really let them down this time around. I bet those who bought the 3080 are not feeling too great about the ti model being released so quickly after they were reassured there wasnt going to be one(not for a while in any case).
Although whats announced and what actually gets to market, are two different things altogether this year.
I'm praying AMD pull it off and get back on top form rolling into 2021 as top dog... I cant wait to have Radeon card in what will be an almost fully AMD system (again).
I feel like 6 more SKU's from AMD would be smart add 3 below the RX6800 with 44/48/52CU's and 3 above the RX6800XT with 88/92/96CU's. The both designs could reduce the infinity fabric to 96MB with the higher end design swapping out the GDDR6 for HBM2 to offset the bandwidth while reducing power and allow for die space for the additional CU's. I think I'd just retain the same bus width or make one 192-bit and the other 320-bit. Probably keeping the bus the same would be easiest and less involved, but either way. I think though with those CU value's and the same board design and infinity cache size it would make it easier to ramp up both. The less than ideal die's could be cut in half and turned into low end models while the better quality ones used for the higher end models.