3 things you should read today: Google cloud shines; Vicarious gets $50M; Qualcomm does AI

We learned yesterday afternoon that Google's cloud business is almost certainly bringing in billions
ARCHITECHT
3 things you should read today: Google cloud shines; Vicarious gets $50M; Qualcomm does AI
By ARCHITECHT • Issue #118
We learned yesterday afternoon that Google’s cloud business is almost certainly bringing in billions of dollars in revenue per year, but exactly how much remains a mystery. Not that it matters much.
The exact details are that Google parent company Alphabet’s “Other” category brought in $3.1 billion during the second quarter of 2017. That’s a 42 percent increase over the nearly $2.2 billion over the same quarter last year. (And, for what it’s worth, Google’s second-quarter capex ticked up nearly 33 percent, to more than $2.8 billion.)
Google CEO Sundar Pichai said on the company’s earnings call that Google scored three times as many $500,000+ cloud computing deals during the second quarter than it did last year. According to a report on CRN, Google Cloud’s expanding customer base has also surprised him:
“GCP continues to experience impressive growth across products, sectors and geographies,” Pichai told investors. “Increasingly with large enterprise customers in regulated sectors.”
Google is winning big customers across diverse use cases, Pichai said in response to an investor asking if there were specific types of workloads gravitating to GCP.
“I would say the breadth of what we’ve seen has really surprised me,” Pichai said of the GCP customer base.
I think the observations about deal size and breadth of customers are much more telling than are the revenue numbers, at least for the moment. 
We spent years following Amazon’s “Other” category searching for clues on AWS revenue, only to realize when the company finally broke it out that while we were close on revenue, most people vastly underestimated the AWS profit margins. (AWS, for what it’s worth, brought in nearly $3.7 billion in revenue, and $890 million in profit, during the first quarter.) And when companies like Microsoft or IBM (or even Google, to a degree) report their cloud earnings, they cover such a broad range of products that it’s difficult to tell how much is from IaaS, SaaS, software, email, and anything else that falls under the cloud umbrella.
If Google Cloud, which spotted AWS and even Azure pretty big leads, is earning even half what they are, I’d say that’s pretty impressive. If it’s doing it at margins comparable with AWS—which for years probably had to invest next to nothing in competitive marketing—that would be even more remarkable. But we won’t really know any of this until Alphabet decides to share it publicly.
However, the biggest question about Google’s cloud computing platform has always been whether enterprises will trust it, not whether it can make money. If, as Pichai suggests, Google really is able to match its technological prowess with a desire, and capability, to win large contracts from large enterprises, that has to give its competitors a lot more to think about. All of a sudden, everyone’s favorite criticisms about Google not being serious about the cloud, or not being capable of working with enterprises, start ringing a little hollow.
There were two other stories that really caught my eye today. Here they are:
  • Qualcomm opens its mobile chip deep learning framework to all (TechCrunch): Qualcomm built a “Neural Processing Engine” for its Snapdragon mobile processor lineup, and today it opened up an SDK for developers to start taking advantage of the additional deep learning horsepower. You can get more details on the SDK here. I suspect we’re seeing the makings of something big happening in the smartphone (and general mobile) space, but I can’t say for sure what it is. However, with Intel, Qualcomm, Nvidia, ARM, Apple, Google, Microsoft and any number of other players all trying to optimize mobile hardware for AI workloads, something has to give.
  • Khosla Ventures leads $50 million investment in Vicarious’ AI tech (VentureBeat): Vicarious is a mysterious company. It first raised money back in 2012 and then in 2014, after deep learning brought AI into the fore, it raised a $40 million round that included a who’s who of tech leaders (including Elon Musk, Mark Zuckerberg and Vinod Khosla) and Ashton Kutcher. Vicarious has always been relatively tight-lipped about the company and its neuroscience-based approach to AI, and was flying under the radar for awhile until it reappeared with some new robotics research recently. Now it’s back in the spotlight with $50 million from Khosla and a focus on robotics.

Sponsor: DigitalOcean
Sponsor: DigitalOcean
Artificial intelligence
HubSpot is in a remarkably competitive space that includes an AI-hungry Salesforce. Bringing in some machine learning expertise, especially a working product, is a good idea.
Really, this is another story on the rise of machine learning inside hedge funds, but the part on Taaffeite Capital Management and its algae-inspired algorithms (which have delivered 21 percent returns) is particularly good.
You also apparently can make money by just investing in robotics and AI companies: “The Global X Robotics and Artificial Intelligence ETF (BOTZ) is up 30 percent this year and the ROBO Global Robotics and Automation Index (ROBO) is up 25 percent.”
Baidu has been up to a lot on the AI front this year. This is a just-deep-enough look at what it has been up to at a macro-level—i.e., driverless cars and smart assistants, rather than research advances.
This is a good piece on the state of conversation AI platforms in China, including Baidu’s DuerOS and Alibaba’s AliGenie. It’s also a good reminder of why comparing U.S. vs. Chinese markets isn’t entirely fair.
This is a really interesting case study from Google Cloud customer Kewpie, which is using unsupervised learning to optimize its QA operations by identifying bad ingredients.
The idea behind this paper is to inform the greater robotics community about advances in deep learning and robotics. I’ll assume it’s good and thorough, but the bigger picture is to remember that different disciplines, business units, etc., need to communicate with each other if real advances are actually going to happen.
arxiv.org  •  Share
New ARCHITECHT Show every Thursday; new AI & Robot Show every Friday!
New ARCHITECHT Show every Thursday; new AI & Robot Show every Friday!
Cloud and infrastructure
RightScale’s new Optima service got quite a bit of pickup today, by promising to help customers optimize cloud spending across providers and business departments. But how this is not yet a solved problem is beyond me.
fortune.com  •  Share
ContainerShip seems like a prime acquisition target. Managed container-orchestration is the right business to be in, but it’s going to be very hard to compete with the big boys here over time.
MongoDB is not nearly as revolutionary as it once seemed, but by all accounts it has improved quite a bit, and large enterprises are embracing it as they try to get more agile and build new applications.
Uber breaks down how it built uChat, its internal chat system. Puppet is still proving its worth for managing server configurations as the companies regularly pushes new code.
If you enjoy the newsletter, please help spread the word via Twitter, or however else you see fit.
If you’re interested in sponsoring the newsletter and/or the ARCHITECHT Show podcast, please drop me a line.
Use Feedly? Get ARCHITECHT via RSS here:
All things data
I’m always curious how new target systems for real-time and edge analytics play inside/outside existing Hadoop-Spark-Kafka-etc. ecosystem. Productizing and packaging the experience is a good idea, assuming it takes advantage of existing infrastructure.
The rise of GPU databases is pretty interesting to me, and here’s another reason they might actually break through—fast, in-depth visualizations.
I know precious little about this space, but Chris Mellor at The Register does. Here are the companies he predicts will rise (Pure and Nutanix among them) and fall over the next 10 years.
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.