AMD Radeon PRO W6800 32GB vs AMD Radeon RX 6800 XT 16GB

The Radeon PRO W6800 is the newer part: RDNA 2, launched in 2021, against the Radeon RX 6800 XT's RDNA 2 from 2020. 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 Radeon PRO W6800 against the Radeon RX 6800 XT: FP32 18 vs 21 TFLOPS, FP16 36 vs 41 TFLOPS. Sparse figures, where the vendor publishes them, appear under each dense number in the performance table.

Memory is 32 GB against 16 GB, fed at 512 GB/s versus 512 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 250 W for the Radeon PRO W6800 and 300 W for the Radeon RX 6800 XT. At FP32 that works out to 0.07 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.