Nvidia DGX-1

Nvidia DGX-1 is a server aimed at accelerating deep learning applications. It features 8 nvidia pascal daughter cards [1] with HBM 2 memory, connected by an NVLink mesh network. [2] It bridges the gap between GPUs and AI accelerators in that the device has specific features specializing it for deep learning workloads. .[3] It delivers 170 teraflops of processing power in a single box,and is capable of training the AlexNet machine vision model in 2 hours. [4]

References

  1. ↑ "nvidia dgx-1" (PDF).
  2. ↑ "inside pascal". Eight GPU hybrid cube mesh architecture with NVLink
  3. ↑ "deep learning supercomputer".
  4. ↑ "DGX-1 deep learning system" (PDF). NVIDIA DGX-1 Delivers 75X Faster Training...Note: Caffe benchmark with AlexNet, training 1.28M images with 90 epochs


This article is issued from Wikipedia - version of the 6/19/2016. The text is available under the Creative Commons Attribution/Share Alike but additional terms may apply for the media files.