NVIDIA GeForce RTX 2080 8GB vs NVIDIA Quadro RTX 6000 24GB

Both GPUs launched in 2018: the GeForce RTX 2080 on Turing and the Quadro RTX 6000 on Turing.

Dense throughput for the GeForce RTX 2080 against the Quadro RTX 6000: FP64 0.3 vs 0.5 TFLOPS, FP32 11 vs 16 TFLOPS, FP16 42 vs 131 TFLOPS. Sparse figures, where the vendor publishes them, appear under each dense number in the performance table.

Memory is 8 GB against 24 GB, fed at 448 GB/s versus 672 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 225 W for the GeForce RTX 2080 and 295 W for the Quadro RTX 6000. At FP32 that works out to 0.05 against 0.06 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.