NVIDIA's mid-range GeForce RTX 3060 could beat flagship RTX 2080 Ti

The last few days have been pretty crazy in the GPU world, with purported specs on NVIDIA's upcoming Ampere-based GeForce RTX 3080 Ti blowing us away -- and then Intel with its "Father of All" GPUs in the Xe-HP. In the middle of all of this, the diamond-in-the-rough was lost: RTX 3060.

4

But don't worry... we're going to dive into that right now. In the leaked specs from a few days ago on the upcoming GA1xx range of Ampere GPUs, there was the lowest end card: the GeForce RTX 3060. The GeForce RTX 3060, at least according to these leaked specs, will be based on the GA104 GPU.

Now, just how powerful is the GA104 GPU (again, in these leaked specs)? Pretty damn powerful. We're looking at more power than the current flagship GeForce RTX 2080 Ti which has a price of over $1200. NVIDIA typically charges far less than $500 for a mid-range GeForce RTX 2060 (or RTX 2060 SUPER) in their Gx104 variant.

If the leaked specs are true, we're to expect up to 2560 CUDA cores based on the Ampere GPU, up to 80 TR Cores, up to 320 Tensor Cores, and GPU boost clocks of up to 2GHz. The GDDR6 on this card will be specced at the faster 16Gbps, and it'll launch in Q1 2021 which is still a while away.

NVIDIA will be using the 7nm node from TSMC for its Ampere GPUs, while AMD is about to strike back and hit the high-end enthusiast graphics card market with RDNA 2-based Navi 2X cards in Q4 2020.

The last we've heard on Ampere is that NVIDIA will be using TSMC's 7nm node for it, while the next-gen Hopper GPU could be making an appearance on Samsung's new 5nm EUV process.

NVIDIA Ampere GA100 (RTX 3080 Ti)

• CUDA cores: up to 8192
• RT Cores: up to 256
• Tensor Cores: up to 1024
• GPU Boost clock: 1750MHz
• GDDR6 speed: 16Gbps
• Launch: Q4 2020

NVIDIA Ampere GA103 (RTX 3080)

• CUDA cores: up to 4608
• RT Cores: up to 144
• Tensor Cores: up to 576
• GPU Boost clock: 2000MHz
• GDDR6 speed: 16Gbps
• Launch: Q3 2020

NVIDIA Ampere GA103 (RTX 3070)

• CUDA cores: up to 3584
• RT Cores: up to 112
• Tensor Cores: up to 448
• GPU Boost clock: 1900MHz
• GDDR6 speed: 16Gbps
• Launch: Q3 2020

NVIDIA Ampere GA104 (RTX 3060)

• CUDA cores: up to 2560
• RT Cores: up to 80
• Tensor Cores: up to 320
• GPU Boost clock: 2000MHz
• GDDR6 speed: 16Gbps
• Launch: Q1 2021

More reading on NVIDIA Ampere:

NVIDIA Ampere GPUs

• GA102 - 84 SMs / 5376 CUDA cores / 12GB GDDR6 / 384-bit bus - 40% faster than RTX 2080 Ti
• GA103 - 60 SMs / 3840 CUDA cores / 10GB GDDR6 / 320-bit bus - 10% faster than RTX 2080 Ti
• GA104 - 48 SMs / 3072 CUDA cores / 8GB GDDR6 / 256-bit bus - 5% slower than RTX 2080 Ti

Older NVIDIA Ampere GA100 Specs

• 8192 CUDA cores @ 2GHz (2.2GHz boost)
• 1024 Tensor Cores
• 130 RT Cores
• 48GB of HBM2e memory @ 1.2GHz
• 300W TDP
• TSMC 7nm+
• 36 TFLOPs peak output

I've already written about rumors that NVIDIA's next-gen Ampere GPU architecture would be up to 75% faster than current-gen GPUs such as the Turing architecture, right after rumors that Ampere would offer 50% more performance at half the power of Turing. This is pretty crazy stuff right there.

Not only that, but we've got some rumored specs on the purported GeForce RTX 3080 and GeForce RTX 3070 graphics cards, which will both be powered by NVIDIA's new Ampere GPU architecture.

We've already heard that Ampere would offer 50% more performance at half the power of Turing, which sent the hairs on my neck standing up. Better yet, you can read about the leaked specs on the purported Ampere-based GeForce RTX 3080 and GeForce RTX 3070 right here.

GA103 (GeForce RTX 3080)

• 10/20GB GDDR6
• 320-bit memory interface
• 60 SMs
• 3480 CUDA cores

GA104 (GeForce RTX 3070)

• 8/16GB GDDR6
• 256-bit memory interface
• 48 SMs
• 3072 CUDA cores

AMD also just revealed its next-gen RDNA 2 / Navi 2X plans, something we referred to as Big Navi until just recently. Those specs are monstrous, but they pale in comparison to what NVIDIA has planned for its flagship HPC destined Ampere GA100 beast.

AMD RDNA 2 / Navi 2X / Big Navi Specs

• 7nm node (confirmed)
• RDNA 2 architecture (confirmed)
• 80 compute units
• 5120 stream processors
• 17.5 TFLOPs compute performance
• Hardware ray tracing support (confirmed)