But does AWS really need VMware?

The big news coming out of VMware's annual VMworld event this week is that the company's promised VMw
But does AWS really need VMware?
By ARCHITECHT • Issue #129
The big news coming out of VMware’s annual VMworld event this week is that the company’s promised VMware-on-AWS product is now available. This is probably very good news for VMware, which needs a viable cloud product after years of confusing cloud strategies that only got fuzzier as a result of the Dell-EMC merger last year. Letting customers run and manage their VMware machines on the Amazon Web Services cloud is an approach that seems to fit that bill.
However, it’s hard to see how this partnership is going to move the needle much for AWS. Look at the companies’ most-recent fiscal years—2016—during which VMware grew about 9 percent to just over $7 billion in revenue, while AWS grew about 45 percent to more than $12.2 billion in revenue. It’s on pace for about $16 billion in revenue in 2017, and an investment analyst at Jefferies predicts AWS will hit $55 billion in revenue by 2022.
All of this without running what will likely be a relatively small number of VMware instances. 
I might be wrong, but I look at this as essentially a larger (Nutanix does less than $1 billion in annual revenue at this point), and less container-focused, version of the deal Google and Nutanix announced in June. There, too, the on-prem partner of the hybrid cloud partnership is delivering workloads from existing customers to the public cloud partner, but I don’t see too much traffic going the other way. AWS and Google collect some additional revenue from partners’ customers, while the partners own the customer relationship and largely dictate the user experience.
I think the more interesting hybrid cloud approaches are the ones where the cloud—or cloud-native—side of the partnership is leading the charge. When and where companies want the benefits of the cloud, the whole idea for most is that they expressly do not want the status quo. Microsoft is able to do this by itself via the Azure Stack appliance, which brings the Azure experience and service inside a company’s data center. Container-orchestration companies want to abstract management, scaling, etc, to layers above the VM altogether, running wherever those machines happen to be hosted. 
There will continue to be a lot of big workloads running inside company data centers. If AWS and Google really want a shot at owning them, they’ll probably need to get their hands (and code) a little dirty by going to where those applications live and showing there’s a better way of doing things.

Sponsor: DigitalOcean
Sponsor: DigitalOcean
Artificial intelligence
To some degree, Intel’s continued relevance in AI could hinge on how widely adopted its Movidius chips are adopted in cameras, drones and other embedded AI applications. That’s where the biggest opportunity is right now, and it’s not inconceivable that “Movidius Inside” is the new “Intel Inside.”
That Musk would raise money for his brain-interface startup is hardly news, but there’s a little controversy here: it looks like the company is seeking a $100 million total investment, but Musk says Neuralink isn’t raising money …
This is some interesting work between Intel, Stanford and the National Energy Research Scientific Computing Center. Baidu has essentially built supercomputers for speech-recognition tasks, so it seems we should be doing it for scientific computing, as well.
The argument here is that servers will require fabrics including GPUs, CPUs and perhaps specialized processors for security. 
This is a take on Y Combinator’s current batch of AI startups, in which the author questions which ones might stand a chance at success. If companies would talk about “machine learning” or even just “data science"—which is probably what many actually do—the discussion might be about the application rather than the algorithms.
Hint: Take everything you heard about “big data” and actually do it this time.
hbr.org  •  Share
This is one of those “big if true” situations. For starters, it would be a black eye for autonomous cars, especially if Tesla’s technology is proven to be unsafe. Further, there’s the irony that maybe this is the AI Elon Musk really should be afraid of.
AI security is going to become a much bigger deal the more we rely on machines to keep us safe. Stealing data is one thing, but messing with the systems that smarten up, say, driverless cars could cost lives.
qz.com  •  Share
This is pretty clever: The tool will, essentially, test an algorithm over and over to see how changes in variables (e.g., race) affect the outcome. I wonder how it works with latent bias even in the data, such as where lots of other data points can act as proxies for race.
gcn.com  •  Share
This is from the thing Google did a while back, where it had people draw objects. I’m not sure how commercially valuable this data is, but it seems like there are some cultural insights lurking inside it.
It’s kind of a meta undertaking—building an AI system can tell if another AI system is producing good dialogue. 
arxiv.org  •  Share
Resource sharing could potentially be a major concern as more organizations attempt to do AI/ML without breaking the bank. These researchers have devised a platform to make that process more efficient and fair.
arxiv.org  •  Share
This is just a good use of computer vision. 
Sponsor: Bonsai
Sponsor: Bonsai
Cloud and infrastructure
Your regular reminder that it’s always worth considering where your data centers are located, how likely a natural disaster will be, and how well the facility is designed to prevent downtime from disaster.
Artificially inflated, as the headline link says? Or so popular Oracle needs to hire 5,000 more people to work on it? Perhaps Oracle just believes if it builds a cloud, they will come.
This is a good article looking at how Microsoft is trying to patch up holes in a long-vulnerable part of its operating system—and that will probably just force hackers to find a new part to exploit.
This is not a bad idea at all, especially as microservices and low-power IoT workloads ramp up. However, it seems unlikely that cloud providers would run with this idea, meaning its commercial viability is probably minimal.
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
This is would be a win for the Riak community, which appears to love the technology and will hopefully be able to foster its continued improvement and stability.
It’s called KSQL and it seems pretty awesome, especially as requirement around real-time everything—from customer service to security—pick up. The big idea is that it continuously queries data as it streams in, rather than looking something up one time.
Confluent might be comprised of former LinkedIn engineers (they created it while there) and contribute a lot of Kafka development, but LinkedIn is still a very big Kafka user. It’s latest contribution to the project it launched is Cruise Control, a system for managing resource levels in Kafka to ensure SLAs.
It’s not the sexiest topic around anymore, but BI on Hadoop is still a problem that lots of companies care about—because they’re doing it. Seems like Dremio might present some new competition here, as well.
This is a pretty cool story, taking place only several years after people started talking about how big data would take over industries like restaurants. However, the flipside here is that the restaurant industry in some ways benefits a lot more from focusing on good service and good food, as some professionals note in the piece.
If you care a lot about huge, distributed graph databases, this is probably something you want to read.
Of course, getting access to it ethically is a whole other discussion. But one of the big battles in data is figuring out where to find the insights you need—and what people are talking about is probably a good place to start.
New ARCHITECHT Show every Thursday; new AI & Robot Show every Friday!
New ARCHITECHT Show every Thursday; new AI & Robot Show every Friday!
Did you enjoy this issue?
ARCHITECHT delivers the most interesting news and information about the business impacts of cloud computing, artificial intelligence, and other trends reshaping enterprise IT. 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.