NVIDIA L40S 48GB vs NVIDIA RTX 6000 Ada 48GB

Both GPUs launched in 2023: the L40S on Ada Lovelace and the RTX 6000 Ada on Ada Lovelace.

Dense throughput for the L40S against the RTX 6000 Ada: FP32 92 vs 91 TFLOPS, FP8 733 vs 729 TFLOPS. Sparse figures, where the vendor publishes them, appear under each dense number in the performance table.

Memory is 48 GB against 48 GB, fed at 864 GB/s versus 960 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 350 W for the L40S and 300 W for the RTX 6000 Ada. At FP32 that works out to 0.26 against 0.30 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.