NVIDIA updating mid-range GeForce RTX 3060 Ti with GDDR6X in October

NVIDIA is reportedly preparing new versions of its mid-range GeForce RTX 3060 Ti and GeForce RTX 3060 graphics cards, launching in October -- which should see the new mid-range Ampere GPUs launching alongside NVIDIA's next-gen GeForce RTX 40 series GPUs.

2

NVIDIA's current-gen GeForce RTX 3060 Ti Founders Edition graphics card

The rumors on the new GeForce RTX 3060 Ti are coming from leaker "MegaSizeGPU" while VideoCardz reports that their source has "now confirmed that at least two of the models are planned, and they should be expected by the end of October".

The new GeForce RTX 3060 Ti will reportedly be getting better VRAM, with NVIDIA rumored to be using GDDR6X memory on the new GeForce RTX 3060 Ti. The current GeForce RTX 3060 Ti has 8GB of GDDR6 memory at 14Gbps bandwidth on a 256-bit memory bus with up to 448GB/sec of memory bandwidth, while the new RTX 3060 Ti will have the same 8GB framebuffer, but with beefed up GDDR6X -- the "X" is big here.

• Read more: NVIDIA GeForce RTX 3060 Ti with GA103-200 GPU spotted, coming soon?

NVIDIA's purported new GeForce RTX 3060 Ti has 8GB of GDDR6X memory that's clocked much higher -- 19Gbps from 14Gbps on the regular RTX 3060 Ti with 8GB GDDR6 memory -- and even on the same 256-bit memory bus the memory bandwidth skyrockets up to 608GB/sec memory bandwidth -- matching the GeForce RTX 3070 Ti.

The slightly higher-end GeForce RTX 3070 Ti has an identical memory configuration to the rumored new GeForce RTX 3060 Ti: 8GB GDDR6X memory at 19Gbps bandwidth on a 256-bit memory bus with up to 608GB/sec of memory bandwidth.

My take: NVIDIA is going to be launching a super-high-end new GeForce RTX 4090 graphics card (or something faster, we could expect a change in the launch lineup this time around) but it might skip out on launching the new GeForce RTX 4080 and GeForce RTX 4070 graphics cards... but why? Because there are so freaking many GeForce RTX 3080, RTX 3080 Ti, RTX 3090, and RTX 3090 Ti graphics cards on the market... too many, so many in fact that NVIDIA is going to probably change its launch plans for Ada Lovelace because of it.

We aren't even taking into consideration the crypto mining market, and the second-hand graphics card market that will be absolutely flooded in the coming weeks leading up to the holidays... right as the next-gen Ada Lovelace cards are launching.

NVIDIA can't be launching a new sub $1000 graphics card in the form of a GeForce RTX 4080 or similar, while there are already sub $1000 custom GeForce RTX 3080 Ti graphics cards for sale brand new. NVIDIA can't launch a new GeForce RTX 4080 that is 30-50% faster than the RTX 3080, at a higher price than the RTX 3080 Ti is right now.

This is placing NVIDIA in a worse spot than AMD as they sell, and thus have far more graphics cards in the supply chain than AMD does. This is going to cause major issues for the Ada Lovelace launch, while NVIDIA is also facing the question: just how good is the flagship RDNA 3-based Navi 31 GPU from AMD?

• Read more: NVIDIA GeForce RTX 4090 latest: 450W AD102 in October, 600W later
• Read more: NVIDIA's next-gen GeForce RTX 4090 rumored to use new GDDR7 memory

AMD's new RDNA 3-based Navi 31 GPU is going to cause NVIDIA to only have a nanosecond to pause and assess the situation, responding with an even higher-end GeForce RTX 4090 Ti graphics card... or if needed, a new TITAN-class Ada Lovelace monster GPU with 48GB of GDDR6X memory... or even next-gen GDDR7 memory that's right around the corner.