It is looking like 2020 is going to be a great year for NVIDIA and its Ampere-based technology. As the fervor for its upcoming consumer line of GPUs continues to build, momentum for its AI solutions is in full swing. NVIDIA has broken no less than 16 records in MLPerf Training v0.7 with its A100 GPUs. This, also, is on top of it being recently reported as the fastest GPU ever in Octabench.
The A100 Tensor Core GPU demonstrated the fastest performance per accelerator on all eight MLPerf benchmarks. For the overall fastest time to the solution at scale, the DGX SuperPOD system, a massive cluster of DGX A100 systems connected with HDR InfiniBand, also set eight new performance milestones. The real winners are customers applying this performance today to transform their businesses faster and more cost-effectively with AI.
As the company mentioned in its post, the real winners are the customers using this record-breaking technology, and it hasn’t taken long for them to line up in adopting the DGX A100. Within days of its release announcement, it was reported how the Atos BullSequana X2415 would feature the new GPU. Shortly after that, AMD announced it would be pairing its data center EPYC CPUs with the A100. For the moment, it seems there’s no stopping the DGX A100 and its Tensor Core technology. These records are all the more impressive, since the GPU is its first processor based on the Ampere architecture. Congratulations NVIDIA!