NVIDIA Titan V 12GB vs NVIDIA V100S PCIe 32GB

The V100S is the newer part: Volta, launched in 2020, against the Titan V's Volta from 2017. Newer architectures typically add lower-precision formats and better throughput per watt, so check the precision rows your workload actually uses.

Dense throughput for the Titan V against the V100S: FP64 7.5 vs 8.2 TFLOPS, FP32 15 vs 16 TFLOPS, FP16 110 vs 130 TFLOPS. Sparse figures, where the vendor publishes them, appear under each dense number in the performance table.

Memory is 12 GB against 32 GB, fed at 653 GB/s versus 1.1 TB/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 250 W for the Titan V and 250 W for the V100S. At FP32 that works out to 0.06 against 0.07 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.