Samuel Cozannet

[Edit 2017-04-20] A careful reader informed me (thanks for that HN user puzzle) that it is no longer required to run in privileged mode to access the GPUs in K8s. I therefore removed a note that previously stated this requirement, and am in the process of updating my Helm charts to remove it as well

A few days ago, someone shared with me a project to run video transcoding jobs in Kubernetes. During her tests, made on a default Kubernetes installation on bare metal servers with 40 cores & 512GB RAM, she allocated 5 full CPU cores to each of the transcoding pods, then scaled up to 6 concurrent tasks

Here we are. After having spent 21min reading how to build a GPU Kubernetes cluster on AWS, 7min on adding EFS storage, you want to get to the real thing, which is actually DO something with it. So today we are going to define, design, deploy and operate a Deep Learning pipeline. So what is

Earlier this week we built a GPU cluster and installed Kubernetes so that we can do some advanced data processing. What is the thing you need next right after you have GPUs? Data. Data. and Data. And technically, if you looked at any of the tutorials for Tensorflow or the recent PaddlePaddle blog posts, you’ll

A few weeks ago I shared a side project about Building a DYI GPU cluster for k8s to play with Kubernetes with a proper ROI vs. AWS g2 instances. This was spectacularly interesting when AWS was lagging behind with old nVidia K20s cards (which are not supported anymore on the latest drivers). But with the

When I talk about Ubuntu and Kubernetes, and how we deploy the latter at Canonical using Juju, the main question I get is: Can you deploy in an existing infrastructure? Often, existing infrastructure means the VPC and/or subnets that I have been allocated to do my work on AWS. What is better than a little

I don’t know if you have ever seen one of the Orange Boxes from Canonical These are really sleek machines. They contain 10 Intel NUCs, plus an 11th one for the management. They are used as a demonstration tool for big software stacks such as OpenStack, Hadoop, and, of course, Kubernetes. They are freely available

This week at the Openstack Developers Summit we are excited to showcase how Canonical with IBM, Mesosphere, Skymind and Data Fellas are working together to make the opportunities of deep learning easier for everyone to access. Deep learning is a completely new way of building applications. These applications, built around neuronet models

Our friends over at Meteorite BI are progressing along their path to bring Saiku Reporting to as many users as possible. One of their core targets is to get it integrated into Saiku 4 Core and onto Canonical’s Juju service modelling platform as soon as possible. Now, they need our help to make this happen. Coming

In the first part of this two part blog we looked at why Canonical believes a new language is needed for modeling modern applications in the cloud. In this second blog we will apply these high-level concepts to build a modular and scalable sentiment analysis application with Juju, using components such as Kafka, ZooKeeper, Storm

Over the last 20 years, applications have moved from a monolithic mainframe perspective to a scale out vision. However, during this time the representation of applications hasn’t changed that much. This has led to major inefficiencies in their design. In the first part of this two part blog, I will explain why Canonical believes that