GeForce RTX 3090 and RTX 3080 And Titan RTX Variants CUDA Cores And Memory Specifications Leak?

The next-gen NVIDIA GeForce RTX 30 Series and their Titan and Ti variants have started to appear in leaks. The Ampere-based successors to Turing-based Graphics Cards haven’t been officially confirmed or launched by NVIDIA. However, the alleged specifications of NVIDIA’s next-generation GeForce RTX 3090, RTX 3080, and Titan RTX variants have been detailed.

NVIDIA GeForce RTX 30 series, based on a brand-new graphics architecture, has been appearing in leaks, rumors, and claims. Now their key driving features which will be critical for Ultra High Definition (UHD) and high frame-rate gaming performance in next-generation AAA titles have allegedly leaked online.

NVIDIA GeForce RTX 3090 Or 3080 Ti Variant Specifications And Features:

The NVIDIA GeForce RTX 3090 or quite possibly the higher-end ‘Ti’ variant of RTX 3080, is by far the fastest Ampere-based gaming Graphics Card from NVIDIA. Interestingly, it is highly likely that the NVIDIA GeForce RTX 3090 could be an internal codename for the RTX 3080 Ti. Whichever the case, the GeForce RTX 3090 will reportedly feature the GA102-300-A1 GPU.

[Image Credit: WCCFTech]
The GA102-300-A1 GPU is believed to feature 5248 CUDA Cores or 82 SMs (Streaming Multiprocessors). Simple math indicates this is a 20 percent increase in CUDA Cores or SMs over the GeForce RTX 2080 Ti. It is quite likely that NVIDIA could pack 12GB of GDDR6X memory that runs on a 384-Bit Memory Bus. This means the memory could have a speed of 21 Gbps and will be able to offer up to 1 TB/s bandwidth. There’s no information on clock speeds and TMU/ROP counts

NVIDIA GeForce RTX 3080 Ampere Graphics Card For High-End Gaming:

The NVIDIA GeForce RTX 3080 will reportedly be featuring the GA102-200-Kx-A1 SKU. This card will feature the same 4352 CUDA cores as the current generation NVIDIA GeForce 2080 Ti and a total of 68 SMs. The card is reportedly going to pack 10 GB GDDR6X memory running at 19 Gbps across a 320-bit bus interface with a bandwidth of 760 GB/s.

[Image Credit: WCCFtech]
The current generation of Turing-based NVIDIA GeForce RTX 2080 graphics cards used TU104, while the next-generation Ampere-based cards will reportedly use the GA102. The most likely reason is to accommodate higher wattage and thermals and still attempt to keep the prices from soaring. If NVIDIA can manage to keep the prices of this upcoming premium graphics card below $500, it would become difficult for AMD to push its premium cards.

NVIDIA 2nd Gen Titan RTX Ampere:

Rumors claim the next-gen TITAN would have 5376 CUDA cores and 24 GB of memory. It is quite strange that the leak deliberately names the card as 2nd-generation Titan. This means the card will utilize the GA102-400-A1 GPU. The GPU will be composed of 84 SMs for a total of 5376 CUDA cores.

The card is reported to feature 24 GB of GDDR6X memory across a 384-bit bus interface, and a memory bandwidth of 816 GB/s. This is remarkably slower than NVIDIA GeForce RTX 3090 possibly due to a slower pin speed of 17 Gbps. Nonetheless, the slight reduction in speed will be covered by enhanced memory compression architecture. Additionally, buyers will have the advantage of double the memory buffer as the RTX 3090.

The NVIDIA GeForce RTX 30 Series of Ampere-based Graphics Cards are expected to be officially launched in the last quarter of the current year. Incidentally, AMD is expected to reveal its next-gen Radeon Graphics Cards based on the Big Navi Architecture during the same time period.

Alap Naik Desai


A B.Tech Plastics (UDCT) and a Windows enthusiast. Optimizing the OS, exploring software, searching and deploying solutions to strange and weird issues is Alap's main interest.