Huawei Atlas 300I Duo 48GB vs Huawei Atlas 300I Pro

The Atlas 300I Duo is the newer part: Da Vinci, launched in 2022, against the Atlas 300I Pro's Da Vinci from 2021. 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 Atlas 300I Duo against the Atlas 300I Pro: FP16 140 vs 70 TFLOPS. Sparse figures, where the vendor publishes them, appear under each dense number in the performance table.

Memory is 48 GB against 24 GB, fed at 408 GB/s versus 205 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.

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.