Posted on

Dihuni Announces NVIDIA A100 SXM4 NVLink Ampere Architecture GPU based Deep Learning Servers

We are pleased to offer new Deep Learning servers based on the latest and most advanced A100 GPUs designed per NVIDIA’s Ampere architecture announced today. Our NVIDIA A100 servers are suitable  for all AI workloads, offering unprecedented compute density, performance, and flexibility in the world’s first 5 petaFLOPS AI system. NVIDIA A100 servers feature the world’s most advanced accelerator, the NVIDIA A100 Tensor Core GPU, enabling enterprises to consolidate training, inference, and analytics into a unified, easy-to-deploy AI infrastructure that includes direct access to NVIDIA AI experts.

Dihuni’s A100 Deep Learning servers are fully customizable and built to order directly by manufacturers including Supermicro, HPE, Dell etc.

To checkout newly released models, please visit this page.

Key features of A100 Deep Learning Servers:

  • Multi-GPU optimized thermal designs for highest performance and reliability
  • Advanced GPU interconnect options for best efficiency and lowest latency
  • Leading Ampere GPU architecture
  • NVIDIA A100 NVLink SXM4 GPUs
  • Designed for AI / Deep Learning, Virtual Desktop, Scientific Research, 3D Rendering Farm, HPC etc
  • Option to install Dihuni’s AI Software Stack including NVIDIA Cuda Toolkit, NVIDIA DIGITS, NVIDIA TensorRT, Keras, NVCaffe, Caffe2, Microsoft Cognitive Toolkit (CNTK), MXNet, PyTorch, TensorFlow, Theano, and Torch Deep Learning Framework Software