NVIDIA GeForce RTX 5090 32GB vs NVIDIA GB200 Grace Blackwell Superchip 372GB

The GeForce RTX 5090 is the newer part: Blackwell, launched in 2025, against the GB200's Blackwell from 2024. 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 GeForce RTX 5090 against the GB200: FP64 1.6 vs 80 TFLOPS, FP32 105 vs 160 TFLOPS, FP16 419 vs 5,000 TFLOPS, FP8 838 vs 10,000 TFLOPS. Sparse figures, where the vendor publishes them, appear under each dense number in the performance table.

Memory is 32 GB against 372 GB, fed at 1.8 TB/s versus 16.0 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.

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.