Deep Learning Blog | Bitfusion

New Deep Learning and Cuda AMIs

Written by maciej | Feb 1, 2016 2:45:01 PM

Recently we’ve been evaluating various deep learning frameworks including Torch, Caffe, and Tensorflow to understand how they perform and scale be it on CPUs or GPUs. While evaluating performance across a vast array of hardware, ranging from singe nodes with low end GPUs to multi-node Titan X and K80 clusters featuring InfiniBand, we found that more often than note we had to build VM or bare-metal images to get up and running. As you can imagine, building the VMs, while not exactly a complex task, can be vastly time consuming particularly when one has to sort out incompatibilities between Nvidia driver versions, Cuda versions, cuDNN versions, and application versions. Long story short, we realized that we would be better off formalizing this process into tested, reliable, and optimized images which can be easily built and updated whenever there are drive updates, application updates, Cuda updates and so on.

We've decided to make three of these images available on AWS via AMIs to save you time and frustration. The three AMIs are as follows:

[row]
[column md="8"]

Bitfusion Ubuntu 14 Cuda 7 GPU (Free Trial)

This AMI comes pre-installed with Ubuntu 14, Nvidia Drivers and the Cuda 7 Toolkit. Designed to allow developers as well as users who utilize applications which rely on GPUs and Cuda to get up and running quickly without wasting time on installation and configuration.

[/column]
[column md="4"]

[/column]
[/row]

[row]
[column md="8"]

Bitfusion Ubuntu 14 Cuda 7 Caffe GPU (Free Trial)

This AMI comes pre-installed with Ubuntu 14, Nvidia Drivers, Cuda 7 Toolkit, cuDNN 3, and Caffe. Several Caffe demos are pre-installed as well which show the utilization of GPUs through the use of cunn and cudnn. Get started with Caffe quickly without worrying about installation and setup.

[/column]
[column md="4"]

[/column]
[/row]

[row]
[column md="8"]

Bitfusion Ubuntu 14 Cuda 7 Torch 7 (Free Trial)

This AMI comes pre-installed with Ubuntu 14, Nvidia Drivers, Cuda 7 Toolkit, cuDNN 3, and Torch 7. Several Torch demos are pre-installed as well which show the utilization of GPUs through the use of cunn and cudnn. Get started with Torch quickly without worrying about installation and setup.

[/column]
[column md="4"]

[/column]
[/row]

These AMIs are in additional to the two AMIs we released previously which were targeted at Accelerating Scientific Computing and Media Processing.

Additionally, we’ll be releasing Boost versions of these AMIs over the next few weeks along with a few more AMIs as well, so keep an eye out or subscribe to our blog below to get notified.