ArchiTECHt Daily: Are containers Microsoft's ace in the hole in the cloud?

Tuesday's announcement that the Kubernetes container-orchestration system is now GA on Microsoft Azur
ArchiTECHt Daily: Are containers Microsoft's ace in the hole in the cloud?
By ARCHITECHT • Issue #22
Tuesday’s announcement that the Kubernetes container-orchestration system is now GA on Microsoft Azure wasn’t exactly a surprise—it has been in beta for months, and the company hired Kubernetes co-creator Brendan Burns from Google in July—but it’s still a big deal. Basically, you have the second place (and, by all accounts, coming-on-strong) cloud provider in Microsoft providing the most-comprehensive offering of container platforms of anybody in the market. 
I’ll be diving deeper into the shifting sands of the cloud computing market on the website this week, but suffice it to say that becoming the market leader in hosting containerized workloads would be a big win for Microsoft. 
It has a real chance, too, especially in terms of managed container-orchestration (aka container-as-a-service / CaaS) offerings. At least according to people with whom I’ve spoken, they’re often none too impressed with Amazon Elastic Container Service. Folks running Kubernetes like Google Container Engine just fine, but that’s a Kubernetes-only service. Meanwhile, Azure supports Kubernetes, DC/OS and Docker Swarm.
While most of the attention is on Kubernetes, I can assure you from firsthand experience (I worked at Mesosphere until recently) that DC/OS has some big users and that Mesosphere has some big customers for its enterprise edition. By giving customers a platform where they can experiment with, or even run both systems as managed services, Microsoft is showing that it knows a thing or two about how this industry works. 
CaaS platforms are also the foundation for many third-party and/or open source serverless/function-based/lambda computing systems, which are catching on today and will only get more popular over the next several years. And, as Box co-founder Sam Ghods explained on the ArchiTECHt Show podcast recently, a technology like Kubernetes, if it’s popular enough, has the opportunity to ramp up IT productivity across the board by becoming a standard platform for which engineers write tools.
Remember when Azure launched and it was much more PaaS-like than IaaS-like? Well, PaaS is giving way to CaaS in a lot of instances (although, Stitch Fix is a big Heroku shop we found out recently), and Microsoft’s early vision is in the process of being justified. Now it just needs to capitalize and hope the head start it gave Amazon Web Services is not insurmountable.

Who's buying cloud computing
Who's buying cloud computing
Around the web: Cloud and infrastructure
Doing machine learning or HPC? You can now rent GPU instances on Google Compute Engine, backed by Nvidia Tesla K80 hardware.
It looks like the company will build another data center in Reno, Nevada, where it already has existing facilities. In fact, so do lots of companies.  •  Share
That gives it 34 in all. But who’s counting?
I’m not sure there’s any particularly novel insight in this comparison of cloud providers, but we all read all of these articles anyway ;-) Spoiler alert: AWS seems like the winner, even if the author doesn’t come out and say it.
Here’s a nice update on the progress of Project Natick, which Microsoft announced last year. Seems like wave power is a bigger deal than all that water for cooling.  
TL;DR: Netflix employee, your devices are like malware sieves. Fix them. Here’s how.
In case you missed that prediction somehow. It’s all about the network.
Want to sponsor this newsletter, or the ArchiTECHt Show podcast? Email for more information.
Around the web: Artificial intelligence
Yet another new architecture designed for AI a la memsristors and neuromorphic chips. But this one is an organic, electrochemical design.
Let the quantum benchmarking begin! Seems like D-Wave should be part of any of these comparisons though, given its commercial availability.
This seems like a case of over-design and latching onto “the next big thing” as the next standard thing. But do I know?
Text and content analysis is a fine application of machine learning, but I think the media industry has bigger problems than this can solve. (In fact, reversing course, I could make the argument that analytics have been bad for media.)
Around the web: All things data
That actually seems like a lot of money considering we’ve seen this type of thing before, and I believe it’s an integrated capability in some analytic services. It’s an important problem, but very tough to solve at scale.
Another untouched headline from The Register. It’s about Gartner’s recent research suggesting companies are struggling with their big data deployments. (Maybe it’s time for vendors to move up the stack.)
Or maybe I should say bull elephantish <rimshot>. It’s always interesting to watch how public market investors are thinking about open source and big data, and Hortonworks is really the only game in town at the moment.
This could have been bad. Good thing it appears to have been resolved before anyone exploited it.
Did you enjoy this issue?
The most interesting news, analysis, blog posts and research in cloud computing, artificial intelligence and software engineering. Delivered daily to your inbox. Curated by Derrick Harris. Check out the Architecht site at
Carefully curated by ARCHITECHT with Revue. If you were forwarded this newsletter and you like it, you can subscribe here. If you don't want these updates anymore, please unsubscribe here.