Blog Post

Joyent co-founder on why AWS will win, OpenStack will flatline and connected cars will (hopefully) be zombie-proof

Stay on Top of Enterprise Technology Trends

Get updates impacting your industry from our GigaOm Research Community
Join the Community!

Jason Hoffman has never been short on opinions. He wasn’t as the co-founder and CTO of cloud computing provider Joyent, he wasn’t as a fixture at our numerous Structure conferences over the years (by the way, Structure Data is coming up in March), and he isn’t in his new role as vice president of corporate strategy and portfolio management at Ericsson(s adr).

He proved the latter during an appearance on our Structure Show podcast this week, discussing everything from the cloud market he used to live and breathe to the connected car market his currently employer is doggedly pursuing. It was one of the most fun interviews we’ve done and possibly the most fun I’ve ever had talking cloud computing. Here are the highlights, but you’ll probably want to listen to the whole thing, which also features a great breakdown of the state of mobile networking.

(And if you want to learn more about Hoffman, including how he created Joyent in part to try curing his mother’s cancer, read this profile that Stacey Higginbotham wrote on him last year.)

Download this episode

Subscribe in iTunes

The Structure Show RSS Feed

What Amazon Web Services does right

“Hats off to AWS, still,” Hoffman said. “It just really, really goes to show that … [paraphrasing ex-Google engineer Steve Yegge] a product will always be replaced by an equivalent or lesser product that’s platformized.”

AWS knows what people want, it’s able to iterate quickly and, most importantly, it’s accessible. “I continue to expect AWS to just really take that … retail and buying experience type capability and proceed to just rip into all the traditional IT players and do pretty well,” Hoffman said.

However, if understanding platforms, users and business were all it took to be the top cloud provider, Microsoft(s msft) might very well be in that same spot. “The basic difference between Microsoft and Amazon internally that I don’t think a lot of people appreciate is Amazon … really has turned themselves into one platform used internally and externally,” Hoffman explained. “… Microsoft internally, like Google internally, is a bunch of islands.”

What Google needs to do

Said Hoffman:

“What Google needs to basically do is sit down and say, ‘One, this is going to be a permanent line of business for us and is not going away,’ which Amazon has been very vocal about saying. The second thing they gotta do is they have to say, ‘Not only is it not going away, we are planning for it to be as large a line of business for us as our advertising business.’ … Number three, that ‘We’re willing to contractually do the things that need to be de done to sort of get the trust of large customers.'”

Oh, and spend tens of millions of dollars over the next few years to ensure it becomes a leader in the space.

Asked whether he thinks it will actually do these things, especially pledging to make cloud as big as advertising, Hoffman joked, “Well, I don’t know. I’m just telling you what they should say.”

Jason Hoffman at Structure 2013. (c) Pinar Ozger
Jason Hoffman at Structure 2013. (c) Pinar Ozger

Why OpenStack might be a lost cause

“The big challenge around OpenStack stuff is that everyone that’s there talking, talking, talking, talking, talking — none of them has ever run infrastructure. So they have no idea what they’re talking about,” he said. “And there’s a few vocal people that do and have done stuff, and they’re just not able to push things through that entire process as rapidly as they should.”

What’s wrong with the process, Hoffman continued, is this: “You have a process that’s like governance, governance, governance, governance, governance. Run VMs, run VMs, run VMs. And what the process needs to be about it accessibility, accessibility, accessibility, accessibility, accessibility, accessibility. Platform features, platform features, platform features, platform features.”

Until someone contributes the pieces of the stack that OpenStack needs to become a real platform, Hoffman doesn’t see it amounting to more than the Unix-based Common Desktop Environment did as a Windows alternative in the 1990s. “It just isn’t gonna happen,” he said.

What ‘cloud’ means to carriers

“Just like a lot of the internet-facing and webscale companies had to come up with a mobile strategy, on the mobile equipment manufacturers side and even the mobile carrier side, we’re now having to come up with what’s called a digital strategy.”

At Ericsson, that entails everything from managing consumer content to dealing with an internet mostly populated by machines to figuring out how cloud computing will affect the company’s own IT strategy.

“It’s somewhat different in that a lot of the cloud conversation that’s been existing within Ericsson to date has largely been around how does one actually virtualize and then make more accessible really the core of carrier networks, so it’s not even meant to be a public cloud service or even to be consumed by enterprise IT users or anything like that,” Hoffman said. “… The conversation to date has been more around, from the radio access network all the way up to the core network, how does that become more ‘cloudy.'”

"I forgot to pay my data bill!" Source: Flickr / Tom Frisch
“I forgot to pay my data bill!” Source: Flickr / Tom Frisch


The most-important challenges for connected cars

Hoffman, who was at CES this week, also shared some of his thoughts from the show and current efforts to connect everything from baseball stadiums to cars. One of the big challenges to all this connectivity is keeping everything that needs to be online online. “[Y]ou can’t, for example, all of a sudden be driving by AT&T Park and everyone’s tweeting and Instagramming during the game and your car stops working,” Hoffman said.

Carriers or car manufacturers also need to figure out a business model that works for consuming all that data. “You’re not gonna go … ‘OK, it’s $800 a month for your car to be connected.’ No, no thanks,” Hoffman joked. “I really need one that’s gonna go on working during a zombie attack.”

6 Responses to “Joyent co-founder on why AWS will win, OpenStack will flatline and connected cars will (hopefully) be zombie-proof”

  1. Joe Specht

    It’s unclear if it is the writer of the article of Hoffman, but there really isn’t even a coherent argument in here, just a rambling of Jason’s opinions with no factual arguments or specific points. He might have well as said I don’t think Openstack is going to succeed because it smells funny.

  2. Tim Wessels

    Well, Mr. Hoffman does lay it out there whether you agree with him or not. OpenStack is complicated and it will get worse as more programs, formerly called projects, are added to the framework. Mr. Hoffman is right, OpenStack is not a platform that you can just go out and use in your data center. The public cloud OpenStack providers like Rackspace and HP have to do a lot of additional work to make OpenStack a platform that is accessible to their customers. Anyone contemplating running a private OpenStack cloud should be prepared to evaluate private cloud OpenStack platforms from the likes of Piston Cloud Computing and Nebula. At global scale the AWS platform is leading the industry, which is no surprise. Only the likes of Google and Microsoft have any chance of giving AWS some serious competition for the delivery of cloud services over a homogeneous platform. All of this is not to say that OpenStack, CloudStack, Joyent and others won’t be able to survive in the shadow of the giants and be commercially successful. There is a market for boutique and regional cloud providers who are able to differentiate their services to attract customers that AWS cannot service with their existing business model.

  3. I agree with Hoffman’s comments on AWS, but have to disagree on OpenStack.

    At CTERA we’re seeing a move from talk to actual deployments. True, there’s a learning curve, and it will take longer for both service providers and enterprises to make OpenStack work, but if history is anything to go by, then an open source platform supported by industry giants will eventually win (see Linux vs. Solaris). OpenStack will be the Android to AWS’s iOS – harder to use, less integrated, but more open and flexible.

    This isn’t to say that AWS’s growth will be curbed. It’s still early days and there’s plenty of room for exponential growth. But it’s naive to think that there’s only one game in town.

  4. uǝǝpɐʎnɓ sıuǝp

    “And there’s a few vocal people that do and have done stuff, and they’re just not able to push things through that entire process as rapidly as they should.”

    Sounds like Jason is talking about Randy Bias.

  5. Robert Scoble

    Fighting words.

    Don’t run infrastructure? Huh? Rackspace Cloud runs on OpenStack.

    Totally disagree with his conclusions but this is the kind of stuff Microsoft execs would say when they were getting beaten in the marketplace. I hated when they did that because everyone could see who was winning in the marketplace.