NVIDIA GeForce RTX 2060 6GB vs NVIDIA GeForce RTX 2080 Super 8GB

Both GPUs launched in 2019: the GeForce RTX 2060 on Turing and the GeForce RTX 2080 Super on Turing.

Dense throughput for the GeForce RTX 2060 against the GeForce RTX 2080 Super: FP64 0.2 vs 0.3 TFLOPS, FP32 6.5 vs 11 TFLOPS, FP16 26 vs 45 TFLOPS. Sparse figures, where the vendor publishes them, appear under each dense number in the performance table.

Memory is 6 GB against 8 GB, fed at 336 GB/s versus 496 GB/s. More memory per GPU means larger models fit before you have to shard across cards, which often matters more than raw TFLOPS for inference.

Power budgets are 160 W for the GeForce RTX 2060 and 250 W for the GeForce RTX 2080 Super. At FP32 that works out to 0.04 against 0.04 TFLOPS per watt.

For LLM training and inference, weight the FP8 and FP16 rows and memory capacity most heavily. For scientific and HPC workloads, the FP64 row is the one to read.