Launch of NVIDIA GeForce RTX 3080 Ti February, RTX 3060 In January 2021

Igor at Igor’s laboratory just leaked the date for NVIDIA’s upcoming RTX 3080 Ti graphics card. This is the most anxiously awaited card of NVIDIA’s RTX 3000 line considering that it has the full 20 GB vRAM buffer (as opposed to the 10 GB buffer in the 3080) and is expected to be the “true” one. flagship, as the 3090 is essentially a replacement for the Titan. While Igor has been very solid in the past, please remember that NVIDIA’s launch schedules are vulnerable for Jensen to change his mind and other proposals and requirements – so take launch dates with salt.

Igor: RTX 3060 will come in 6GB and 12GB flavors.

According to Igor, the RTX 3080 Ti will land in February 2021 while the RTX 3060 will land early in January 2021. This makes sense, as the RTX 3060 is the budget king for NVIDIA’s lineup and should be the one users will want to buy with large volumes. That said, we’re not sure about the state of NVIDIA’s supply lines, as a few weeks ago it wasn’t visible when the company would be able to reach demand.

Even more interesting, Igor mentions that the RTX 3060 will come in 6GB and 12GB flavors. If this is true, then it would seriously devalue the value of a 10GB RTX 3080. I was complaining about the 10GB buffer right away – because it’s certainly not enough for a 2020 line of games based on active stream. Who knows, NVIDIA might decide to give the RTX 3080 the 20GB treatment even after they have enough backup GDDR6 memory.

Recap: The RTX 3080 Ti 20GB

The NVIDIA RTX 3080 Ti was originally supposed to be based on the PG132 SKU20 (although NVIDIA decided to rename it) and has been in the pipeline for quite some time. Although the correct CUDA core calculation is not yet confirmed, it will have twice the memory compared to the vanilla RTX 3080. To be clear, even the nomenclature is not confirmed, as the SUPER suffix could appear anywhere in the name if Jensen would think it so.
We highly recommend users (who want to buy an NVIDIA card) to choose the RTX 3090 (if they can afford it) or wait for the RTX 3080 Ti / SUPER 20 GB.

With this amount of shadow power, you’ll want to have a free memory buffer. According to Videocardz, the card will land in January 2021. Considering the fact that killer programs like Microsoft Flight Simulator 2020 are already taking away all the vRAM the 2080 Ti had to offer, 10 GB will feel like me in a year or two. New-gen consoles are already driving games with an active stream of big worlds, so vRAM is starting to become much more crucial than before. With AMD struggling mainstream card with 16GB of vRAM, NVIDIA had to respond with its own high-end vRAM card.

According to Kopite, the RTX 3080 Ti will be based on the GA102 and will have specifications between RTX 3080 and 3090. The correct chip nomenclature is GA102-250-A1 and will have a 384-bit bus with GDDR6X memory. The bus size means NVIDIA will use a 12 GB buffer or 24 GB. Considering that Microsoft Flight Simulator 2020 is already blocked by the 11 GB RAM on the RTX 2080 Ti, it would be disappointing to see NVIDIA send another powerful GPU with a small memory buffer. With RTX IO and active streaming (Unreal Engine demo for next-generation consoles) becoming an issue in the next year or so, every bit of a buffer will help in this paradigm shift.

It’s also not clear at this point how the revelation of this new GPU matches the rumors about NVIDIA planning a move back to TSMC. However, it does believe the belief that the RTX 3000 series, at least for now, remains at the Samsung 8nm process. The company initially faced less than ideal returns and supply limits at launch, but those are expected to improve significantly as we enter the new year.

This will be a crazy power card with 34 TFLOP powers. That said, the current API, driver, and application infrastructure cannot fully utilize this raw power. This is the actual reason that NVIDIA’s insanely powerful cards cannot be scaled in line with TLFOPs. The company has basically created cards that are ahead of their time compared to the surrounding software ecosystem.

The difficult proof of this lies in the fact that the RTX 3000 series experienced nonlinear positive scaling while lowering the stack. RTX 3070, which has slightly more cores than the RTX 2080 Ti, surpasses the previous flagship – stopping any rumors or accusations about the CUDA cores in the Ampere series not as strong as the Turing series or misleading faults in architectural design with missing INT performance.

AMD’s big Navi series falls later today and it remains to be seen how the GPU market will shape as we enter the holiday season. NVIDIA’s pricing is excellent, but it needs to work with developers to fix the API and driver stacks to properly take advantage of the raw performance offered by the Ampere series (excellent high-end wine) and meanwhile AMD will create budget-friendly business cards with that, which appears to be ample supply. They have also taken many steps to ensure that the bot scalping that occurred with NVIDIA does not occur at their launch of the Radeon RX 6000 series.

NVIDIA GeForce RTX 30 Series ‘Amp’ Graphics Card Specifications:

Graphic Card Name NVIDIA GeForce RTX 3050 NVIDIA GeForce RTX 3050 Ti NVIDIA GeForce RTX 3060 NVIDIA GeForce RTX 3060 Ti NVIDIA GeForce RTX 3070 NVIDIA GeForce RTX 3070 Ti? NVIDIA GeForce RTX 3080 NVIDIA GeForce RTX 3080 Ti? NVIDIA GeForce RTX 3090
GPU name Ampere GA107 Ampere GA106? Ampere GA106? Ampere GA104-200 Ampere GA104-300 Ampere GA102-150 Ampere GA102-200 Ampere GA102-250 Ampere GA102-300
Process Node Samsung 8nm Samsung 8nm Samsung 8nm Samsung 8nm Samsung 8nm Samsung 8nm Samsung 8nm Samsung 8nm Samsung 8nm
Die Die TBA TBA TBA 395.2mm2 395.2mm2 628.4mm2 628.4mm2 628.4mm2 628.4mm2
Transistors TBA TBA TBA 17.4 Billion 17.4 Billion 28 Billion 28 Billion 28 Billion 28 Billion
CUDA Cores 2304 3584 3840 4864 5888 7424 8704 10496 10496
TMU / ROP TBA TBA TBA 152/80 184/96 232/80 272/96 328/112 328/112
Tensor / RT Cores TBA TBA TBA 152/38 184/46 232/58 272/68 328/82 328/82
Basic Clock TBA TBA TBA 1410 MHz 1500 MHz TBA 1440 MHz TBA 1400 MHz
Accelerate Clock TBA TBA TBA 1665 MHz 1730 MHz TBA 1710 MHz TBA 1700 MHz
FP32 Computing TBA TBA TBA 16.2 TFLOPs 20 TFLOPs TBA 30 TFLOPs TBA 36 TFLOPs
RT TFLOPs TBA TBA TBA 32.4 TFLOPs 40 TFLOPs TBA 58 TFLOPs TBA 69 TFLOPs
Tensor-TOPs TBA TBA TBA TBA 163 SUPPORTS TBA 238 SUPPORTS TBA 285 SUPPORTS
Memory Capacity 4 GB GDDR6? 6 GB GDDR6? 6 GB GDDR6? 8 GB GDDR6 8 GB GDDR6 10 GB GDDR6X? 10 GB GDDR6X 20 GB GDDR6X 24 GB GDDR6X
Memory Bus 128-bit 192-bit? 192-bit? 256-bit 256-bit 320-bit 320-bit 320-bit 384-bit
Memory Speed TBA TBA TBA 14 Gbps 14 Gbps TBA 19 Gbps 19 Gbps 19.5 Gbps
Bandwidth TBA TBA TBA 448 Gbps 448 Gbps TBA 760 Gbps 760 Gbps 936 Gbps
TGP 90W? TBA TBA 180W? 220W 320W? 320W 320W 350W
Price (MSRP / FE) 149 USD? $ 199? $ 299? $ 399 US? $ 499 US US $ 599? $ 699 US $ 899 US? $ 1499 US
Launch (Availability) 2021? 2021? 2021? November 2020? October 29th Q4 2020? September 17th January 2021? September 24th

Source