LeaderGPU® introduces new GPU servers with the NVIDIA® A100

The NVIDIA® A100, one of the most powerful GPUs for solving AI and HPC tasks, is now available to LeaderGPU® customers.

The Tensor Cores in the NVIDIA® A100 with Tensor Float (TF32) support significantly improve performance – up to 20 times more than NVIDIA® Volta™. The A100 includes a ton of inference optimisation capabilities. Using this GPU, you can significantly speed up various calculations: from FP32 to INT4.

The latest generation of A100 40 GB memory helps accelerate the processing of the largest and most time-consuming models and datasets.

Current Configuration Options with A100

You can purchase GPU servers with A100 in LeaderGPU® in the following configuration:

• 8 x A100, 320 GB GPU memory, 2 x Xeon® Gold 6226R

Servers are available with monthly, weekly, day and minute billing.

Subscribe to our updates to stay informed about the latest GPU configurations!