ARCHITECHT Daily: What should we make of the new Docker CEO?

Docker got a new CEO on Tuesday, and lots of folks—myself included—are wondering what's up. Will he t
ARCHITECHT
ARCHITECHT Daily: What should we make of the new Docker CEO?
By ARCHITECHT • Issue #68
Docker got a new CEO on Tuesday, and lots of folks—myself included—are wondering what’s up. Will he try to sell the company? Take it public? Just get it making some real money? 
What I do know is that rumors had been swirling for quite a while that former CEO Ben Golub was on his way out, so many industry watchers were just waiting for the other shoe to drop. While Docker containers have achieved seeming ubiquity, many will tell you the company has not been generating adequate revenue to justify its lofty valuation. 
Supporting hundreds of employees in Silicon Valley is expensive, especially if it includes a skilled enterprise sales team. Docker actually appears to have a reasonable number of customers, but it takes a lot of smallish support deals to add up to something meaningful. Furthermore, as I’ve written here before, making money in the container space is doubly difficult because there’s a built-in expectation of open source and a natural inclination to focus on developers. 
You know who appears to be making a killing off of Docker? Red Hat. Why? Because Docker is doing, and open sourcing, much of the foundational work—not to mention driving mindshare—and a large company like Red Hat can swoop in with its Kubernetes-based OpenShift technology and offer an entire platform experience. Backed and supported by Red Hat, no less.
I linked to this Red Hat blog post yesterday, but let me call out this section:
[Q3, FY 2017] “Also of note, we closed our second OpenShift deal over $10 million and another OpenShift deal over $5 million. And significantly, we actually had over 50 OpenShift deals alone that were six or seven figures, so really strong traction.“
[Q4, FY 2017] “with our largest deals in Q4 … approximately one-third had an OpenShift container platform component.
This isn’t to say Red Hat is going to own the platform-as-a-service/container-as-a-service/container orchestration/whatever-you-want-to-call-it space forever, but Red Hat’s present (and relative) success does highlight why there’s so much activity at that layer of the container stack. When Docker realized it needed to monetize its open source technology beyond support, it walked into a higher-level space crowded by startups and large vendors alike.
Beyond revenue and product strategy—and there’s a lot more to say on that—an underappreciated issue, I think, is the level of controversy and animosity surrounding Docker. Not having been inside the company or even part of the Docker developer community, it’s hard to know for sure what’s going on. But if you look at how Docker arguably has mismanaged everything from community messaging to harassment claims, or if you read its Glassdoor reviews or the comments on Hacker News posts about the company, it’s hard to not leave with a sense that something is amiss.
Granted, no company is perfect, but the spotlight has a tendency to, well, shine a light on imperfections. You could argue that Docker has had to grow up faster than maybe it should have, but that doesn’t change the fact that people are paying attention and forming opinions.
As for Docker’s new CEO, Steve Singh, all accounts (and I acknowledge that I haven’t done a lot of investigating) are that he is a great leader and a great person to work for. (There was this “glitch” on his résumé, but that didn’t appear to materially affect anything.) Some people suggest he was brought in to try and sell Docker, but I’m not sold on that argument. Docker still has a lot of upside.
Yes, Singh did eventually sell his previous company, Concur, to SAP for billions, but he did so only after 20 years and an IPO. Also, I would bet that Docker has fielded its fair share of acquisition offers already, and might even have engaged in some discussions about it. Maybe Docker couldn’t get the terms it wanted and the board thinks Singh is the guy to get it in shape so it can get a good exit. On the other hand, if the pressure really is on to sell, there almost certainly have been opportunities.
[WARNING: Mixed metaphors ahead.] Docker has had an outsized impact during its short existence, and now it’s time to grow into the business suit it’s been forced to wear. Maybe, just maybe, Singh’s mission is to right the Docker ship and turn it into the enterprise IT giant that many people expect it to be. 
Further reading:

Sponsor: Cloudera
Artificial intelligence
Bonsai says it has hit upon a way to help companies train their own AI models, without making them understand AI models. Learn more in this ARCHITECHT Show interview with CEO Mark Hammond
bons.ai  •  Share
Your weekly reminder that AI models will eventually run on many devices, which is why there’s a mad rush to invest in AI-optimized hardware.
On the one hand, stories about AI advancing science and medicine are becoming common. On the other hand, it’s hard not to be optimistic when you look at them as a whole.
phys.org  •  Share
It works vs. how it’s working is the new correlation vs. causation. From the article: “Such opacity’s okay in a game-playing machine but not in a self-driving car.”
The Chinese tech giant wants to be an AI giant, too. As I said yesterday, it has its work cut out for it. But here are some arguments for its ambitions.
I really don’t know what he’s talking about. But I think Mark will be OK even if/when he’s proven wrong on this investment.
Bees, as you might know, are extremely important to our world. This kind of ingenuity is inspiring, and the fact that they can now do it is even more amazing.
Meet PennAI, a GUI-based tool for non-experts to run AI and ML experiments. Maybe this will catch on in academia, but I’m pretty sure the cloud providers will own this space commercially.
arxiv.org  •  Share
This could be a remarkably useful paper for folks in the auto industry to read. It highlights today’s tools and products, and how they apply to the automotive industry.
arxiv.org  •  Share
Listen the the ARCHITECHT Show podcast. New episodes every Thursday!
Cloud and infrastructure
At this point, is it safe to assume AMD won’t be the company to challenge Intel’s dominance in the data center?
Vector, a company formed ex-execs from many high-profile places, including SpaceX, is building an OS for satellites. Whether or not they run XenServer (the point of this news), there could be a good business in letting companies deploy satellite applications directly on the satellites.
The author of this post suggests that the cloud market does not revolve around Amazon Web Services, but that it’s just the Jupiter and Saturn of the space. Some might argue it’s the sun.
The author thinks it will; I think it will not. (1) Betting that serverless kills private clouds means betting that serverless explodes in adoption very quickly. (2) Containers will evolve the private cloud, including by enabling serverless computing.
This makes some sense. It would also make sense to believe these workloads can run in micro data centers managed by AWS, Google and Microsoft.
By rolling out its first AWS instances on the continent. That could have a meaningful impact on performance and/or compliance if you’re running Atlassian-managed JIRA.
Interesting paper on the existing, and possible, options for managing all those IoT devices and infrastructure.
arxiv.org  •  Share
Media partner: GeekWire
All things data
When you find a new source of data that so valuable, and find ways to effectively capture it, money will follow. Orbital is one of several high-profile companies in this space.
This is a sponsored post (so acknowledge any vendor bias), but it raises some good points about why big data is still so hard, and some near- and long-term approaches to solving it.
You can hate on Uber all day, but the technological choices it makes to run its platform are still worth absorbing. Here’s a post on how Kafka, Cassandra and other tools help it move important data in real time.
Another sponsored post, but it’s addressing the promising intersection of big data and medicine. UCSF has already done some impressive work in this space, and it’s investing in even more.
We know who won the Booz Allen Hamilton competition (hosted on Kaggle) to improve cancer detection, but we’ll have to wait until Nvidia’s upcoming conference to hear how they did it.
Did you enjoy this issue?
ARCHITECHT
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 https://architecht.io
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.