Author: Barb Darrow

Surprised? Cloud providers bring in $2B in first quarter
Given all the hoopla surrounding cloud computing it, it’s not surprising that revenue from cloud services is growing. Total worldwide revenue from top Infrastructure-as-a-Service and Platform-as-a-Service players hit $2 billion for the first quarter of 2013. That’s up a healthy 56 percent from the year ago period, according to Synergy Research Group.

These numbers factor in AWS and Amazon’s Elastic Beanstalk on the PaaS side; Google App Engine; IBM’s SmartCloud and SmartCloud Application Services; Microsoft Azure and Cloud Services for Windows Azure; Salesforce.com’s Heroku and Force.com (but not its bigger Software-as-a-Service applications business.)

Key takeaways:
- Amazon Web Services remains alone atop the heap with 27 percent of total IaaS and PaaS revenue in Q1, up from 24.7 percent for the year-ago period.
- Salesforce.com is second largest but it’s revenue share decline year over year to 6.8 percent from 7.9 percent.
- North America accounts for more than half the worldwide IaaS/PaaS revenue.
These figures won’t surprise folks who already see Amazon Web Services at a $2 billion-a-year revenue rate. Morgan Stanley thinks AWS alone will hit $24 billion in revenue in the next decade. While other players — Microsoft, IBM, Google and Fujitsu all saw slight growth. Some of the major telcos – AT&T, NTT and Verizon — were off slightly year over year.

These numbers beg the question of whether anyone can catch Amazon. Google just made its Google Compute Engine IaaS generally available, and Microsoft launched its Azure IaaS capabilities in April. Those well-funded entries along with VMware’s hybrid cloud service coming online, will pose more competition for the AWS monolith.

AWS boosters say the company’s 6-year-head start makes it invincible, but many workloads have yet to migrate: IDC estimates that just 5 percent of total IT spend is now in the cloud. That leaves lots of upside opportunity for the companies who delivers the best, most flexible and cost efficient services.

Who that might be is up in the air, but one place you can hear all about the top candidates will be at GigaOM’s Structure where a list of cloud powerhouses including VMware CEO Pat Gelsinger, Amazon CTO Werner Vogels and Microsoft Server and Tools group president Satya Nadella will all be on hand to talk up their companies’ cloud strategies.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.
May 30, 2013
Microsoft caves on Windows Start button. Kinda
Microsoft Windows 8.1 will offer users a more familiar desktop mode featuring — you guessed it — a Start button, albeit one that looks different from the old standby. Previews of the new operating system upgrade have been promised for attendees of Microsoft’s Build Conference in June.

In a blog post, Antoine Leblond, corporate VP, provided some detail around the 8.1 release. He wrote:

You can even choose your desktop background as your Start screen background, creating a greater sense of unity and familiarity. And the Start screen in Windows 8.1 features a variety of tile sizes including a new large and new small tile, so you can organize your Start screen exactly the way you want it. It’s also even easier to name groups and rearrange tiles. You can now select multiple apps all at once, resize them, uninstall them, or rearrange them. We also found people were accidentally moving tiles on their Start screen so in Windows 8.1, you press and hold (or right click) to move things around.

Windows 8 debuted last fall to mixed reviews. One problem, as our Kevin Tofel reported, was that Windows 8 and its new tiled Metro interface was trying to bridge tablet and PC worlds and that was maybe a bridge too far. But many criticized just how different it looked and that’s a problem for a franchise with hundreds of millions of users.

Microsoft is between a rock and a hard place. It wants to embrace new form factors with an OS optimized for them but also needs to placate millions of existing — and change resistant — users. But some question whether Windows 8 underperformed (although Microsoft says it sold 100 million copies) because users don’t like it or because users just aren’t buying as many PCs.

It’s hard to overstate how big a deal the Windows Start Button is to users. Some of us old timers remember when Microsoft licensed the Rolling Stones “Start Me Up” to promote this very important feature of Windows 95. And on that note, here’s a good excuse to revisit that era:

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.
May 30, 2013
With 1M new downloads in 3 months, Node.js momentum just keeps on keeping on
It’s not surprising that Strongloop would tout the traction Node.js is getting in the market – the startup was founded a few months ago to bring commercially supported versions of the language to Red Hat Linux, MacOS, Ubuntu and Windows.

But even more neutral observers give the server-side JavaScript framework its due. After all, the four-year-old Node.js is great for writing high-performance servers that need to handle APIs and fast data ingress and egress.

Here’s a sampling of Strongloop’s new fun facts about Node.js:
- There have been more than 1 million downloads of the latest V.0.10 release in three months.
- Big name users include Dell, General Motors, Dow Jones, Walmart, Yahoo and Airbnb.
- Node is the second-most popular project on Github.
- From 2011 till now, Indeed.com job postings for Node.js skills soared 22,500 (!) percent.
For another, earlier take on why developers flock to Node, check out this Stacey Higginbotham post from 2011.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.
May 30, 2013
Killer cloud: report says Amazon Web Services threatens all IT incumbents
Amazon Web Services faces growing competition from a dozen or more legacy name-brand IT giants. But instead of taking a hit, it poses a bigger-than-ever threat to the those vendors — all of which are building their own competitive clouds, according to new Morgan Stanley research. Oh, and the researchers project that AWS will hit $24 billion in revenue by 2022. Amazon doesn’t break out AWS revenue, but most pundits figure it passed the $2 billion-a-year mark about a year ago.

The fact that AWS has a huge lead in cloud over the rest of the world is not news to anyone who’s been watching, but these projections could be a wake up call to investors who think tech incumbents — companies like IBM, Microsoft, HP, VMware, Red Hat as well as every telco and hosting provider — can challenge Amazon in cloud computing.

“Applying retail economics to the delivery of technology services well positions Amazon Web Services [to be] a Top 5 vendor within the $152 TAM [total addressable market,] ” according to Morgan Stanley analysts Scott Devitt, Keith Weiss and team.

Nobody’s immune

The move to cloud computing means fewer companies will buy huge numbers of servers and storage arrays for their own use. Over the next 5 years, Morgan Stanley’s expects that 3 percent to 17 percent of current spending could be sucked up by cloud-based IT service providers. AWS represents a key risk for infrastructure vendors EMC, Brocade, NetApp, VMware and Qlogic, in particular, according to the report.

Other key takeaways:
- “We expect on-premise server growth to remain negative long-term on the back of smaller footprints post the adoption of server virtualization combined with new workloads moving to the cloud. Partially offsetting the decline is 20% growth in servers shipped to cloud providers, though some of the demand is fulfilled by whitebox makers like Quanta and Wistron.
- Storage market at risk of decelerating growth that isn’t fully baked into expectations (unlike servers which already declined in 2012). We expect 0-5% storage revenue growth going forward, down from 5-10% historically. EMC and NetApp likely gain share from server vendors, like IBM. We downgrade BRCD to UW, given over 30% of revenue derived from server OEMs.”
AWS as enterprise software power

And then there is enterprise software where Amazon threatens VMware and Red Hat in the virtualization market. And, as we’ve reported, Amazon is pushing hard for enterprise workloads with its DynamoDB NoSQL database and RedShift data warehousing. Those AWS efforts represent a long-term threat to Oracle, SAP and Microsoft. In content delivery, where Amazon’s CloudFront is a factor, Akamai faces a long-term threat.

Morgan Stanley isn’t the first analyst firm to up the ante on AWS expectations. In January Macquarie Capital projected that AWS would account for $38 billion of an overall $71 billion cloud services market by 2015. If you don’t like Morgan Stanley’s take on AWS, hold on, there are bound to be others.

The growth of Amazon’s public cloud infrastructure and its push beyond startups into the enterprise, will doubtless come up at GigaOM’s Structure event where Amazon CTO Werner Vogels will speak.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.
May 29, 2013
Say hello to Netflix Conformity Monkey
We knew Netflix was going to keep churning out “monkeys” to help us with our cloud computing deployments because it told us so. So, meet Conformity Monkey, the latest member of the company’s Simian Army to be made available to the public.

The latest monkey, like its cousins, runs in the Amazon Web Services (AWS) cloud where it looks for rogue instances — or at least instances that aren’t behaving according to predefined rules for best practices.

According to the Netflix blog announcing the debut:

Conformity Monkey determines whether an instance is nonconforming by applying a set of rules on it. If any of the rules determines that the instance is not conforming, the monkey sends an email notification to the owner of the instance. We provide a collection of conformity rules in the open sourced version that are currently used at Netflix and believed general enough to be used by most users. The design of Conformity Monkey also makes it simple to customize rules or to add new ones.

Netflix has built a reputation for wringing the best out of Amazon’s public cloud infrastructure and it’s trying to entrench the open-sourced tools its used there to third parties.

Conformity Monkey is just the latest of what looks to be several monkeys to come.

Photo courtesy of Flickr user anneheathen

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.
May 29, 2013
Formula 1 racing changes pose big data challenge
You’d be hard pressed to find a better use case for big data number crunching than Formula 1 racing, where the cars blasting around the track are largely designed, simulated and built on computer screens. There’s a reason for that: as expensive as all that compute fire power is, it’s still cheaper than doing it all manually. Every day the real team spends testing a car out on track costs $400,000 to $450,000, said Patrick Louis, CEO of the Lotus F1 team.

For that reason, with its use of advanced computational flow dynamics (CFD) and CAD/CAM operations, Formula 1 racing is a demanding test case for compute and storage infrastructure. When the real cars do hit the track, each vehicle runs 240 sensors which generate 25 megabytes of data per lap driven. That data is uploaded via satellite link to the factory — the engine data is split from the chassis data and each stream is analysed for performance and wear and tear.

And now, the Formula 1 racing circuit is navigating a massive mandated change — from V-8 engines with 2.4-liter displacement to more gas-efficient V-6 engines with 1.6 liter displacement. And they have to do that with no loss of speed. That means there are plenty of design challenges ahead, Louis acknowledged.

The upcoming Lotus F-1 car will look like this — complete with EMC, Microsoft and Symantec decals.

F-1 design draws on big data, big time

As a friend (thanks Rochelle!) who is a huge F-1 fan explained to me, the race’s governing body (the FIA) lays out detailed rules about how cars can be built. It mandates, for example, that a flat plank on the underside of the carriage must be X centimeters above the ground and no car can benefit from “moveable aerodynamics.”

Each car maker — companies including Lotus, Ferrari and McLaren – interprets those rules for its own design and spends what some estimate to be hundreds of millions developing cars for each season. Yikes. Each car — an intricate combination of 85 percent carbon fiber composites and 15 percent metal — can cost upward of $2 million.

EMC jumps aboard Lotus F-1 bandwagon

But while FIA rules change annually, the whole deck gets reshuffled this year in an effort to cut energy costs, Louis said. That means new engines and new transmissions. EMC put me on the phone with Louis because Lotus will be using an array of EMC gear, including Vblocks (actually bundles of EMC storage, VMware virtualization and Cisco servers) for its back-office functions that run on Microsoft Dynamics AX.

Lotus is building two projects running in parallel, Louis said. One is a factory where one section is dedicated to the aforementioned Microsoft Dynamics business functions and the second for all the Catia CAD/CAM design applications. The second is tasked with equipping the racing team — which obviously travels from track to track — with the same software and content.

Louis wouldn’t detail what hardware is being displaced but it’s safe to say it’s non-EMC servers and storage. EMC’s CMO Jeremy Burton (who likes to race Mazda Miatas in his spare time) said aspects of the Lotus deal could also include EMC’s Atmos for storage and managing content; Syncplicity to sync and share files across sites; and Data Domain for backup and recovery.

The massive computational fluid dynamics (CFD) applications that simulate how airflows will impact speed and performance, and how the hot exhaust of the car can create more downforce, are not part of this deal, Louis said.

This is a big, high-profile multi-faceted job — one that any tech vendor would kill to be part of. So brace yourself for a wave of promotions from EMC, Microsoft, Symantec (judging from the decals plastered on the car above) and whoever else is involved.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.
May 29, 2013
Amazon spreads net wider by federating Facebook and Google (and AWS) identities
Good news for developers who use Amazon Web Services and want to make those apps available to millions of Facebook and Google users: Amazon Web Services Identity Access Management (IAM) can now “federate” Google and Facebook user identities. Oh, and it also supports AWS new Login With Amazon feature which promises that companies can “securely connect your websites and apps with millions of Amazon.com customers.”

Amazon announced identity federation for enterprise users two years ago. That let businesses grant their own employees access to AWS resources based on the users’ current corporate identity management systems. But this new federation capability spreads the net wider. This federation will let developers authenticate a user with her existing Amazon, Google or Facebook credentials, which then give her access to specific AWS resources using her existing IAM roles.

In his AWS blog post announcing the news, Jeff Wierer, IAM principal product manager, explained a basic use case:
“Imagine you’re developing a mobile app that uses the new Login with Amazon service for authentication, and part of the app’s functionality allows end users to upload an image file as their personal avatar. Behind the scenes, you want to store those images as objects in one of your S3 buckets. To enable this, you need to configure a role that is used to delegate access to users of your app. Roles are configured in two parts:
1. A trust policy that specifies a trusted entity (principal)—that is, who can assume the role. In this case, the trusted entity is any authenticated Amazon.com user.
2. An access policy with permissions that specify what the user can do.”
With services like this one, Amazon continues to push its cloud services as the platform of choice for developers at startups and big companies alike as more public competitors come online.

Photo courtesy of Flickr user jaaro

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.
May 29, 2013
Pure Storage nets new cash from In-Q-Tel to push its “flash for all” effort
Flash storage dynamo Pure Storage has snagged new funding — the amount is undisclosed — from In-Q-Tel, the VC affiliate of the CIA and other security agencies. The fresh cash comes atop a $40 million D round closed last August, which brought the total tally at the time to a robust $95 million.

Matt Kixmoeller, VP of products for Pure Storage.

In-Q-Tel backing should make it easier to get traction in defense and security-related venues which view In-Q-Tel backing as a sort of seal of approval, said Pure Storage co-founder and CEO Scott Dietzen. “The immediate value [of the In-Q-Tel relationship] is in the procurement and sourcing agreements with these agencies,” he added.

The Mountain View, Calif.-based company also unveiled the third major release of its core storage product that adds data-at-rest, 256-bit AES encryption that ensures that customer data is encrypted all the time — even if an array has to be taken out of service for transport or repair,

“We are software guys who focus on value-add deduplication, and encryption — we bake them right in and don’t even let them be turned off,” said Matt Kixmoeller, VP of product marketing and management. “Our hardware-oriented competitors add these things after the fact and then see massive performance degradation as a result.”

Pure’s flash-for-all motto is ambitious — solid-state flash memory is seen as much pricier than traditional disk drives and tape. But Pure executives said they’re serious about bringing the cost of flash storage down at least to spinning disk levels.

“If you take traditional tier 1 drives, you generally pay $3 to $5 per gig, but as you provision it and do things to make it useful, the end price is more like $5 to $10 per gig. We can sell flash at about $5 per useable gig,” Dietzen said.

Dan Iacano, IDC’s research director of storage systems, said Pure’s pitch has some merit.”You can probably still get tier 1 disks for a little it lower, but if you look at the all-in baked in price [of the disk and all the software that makes it truly useful] Pure Storage close,” he noted.

Pure Storage definitely generates buzz but it also faces some talented competitors ranging from Violin Memory and Whiptail to Texas Memory, which IBM bought last year to boost its stake in the super-heated flash memory market. The race to flash-enable the world is going to be fun to watch.

Don’t miss Dietzen and Paula Long, who cofounded EqualLogic and now DataGravity, talk at GigaOM’s Structure event about the role storage plays in our highly distributed world.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.
May 29, 2013
Are Microsoft shareholders mad as hell? One analyst thinks they’re ready to force change
Rick Sherlund, Nomura Securities’ software analyst, clearly thinks something’s up with Microsoft. He is a long-time, respected Microsoft watcher — first at Goldman Sachs and now at Nomura — so when he puts out a research note saying something’s new, even if he’s a little coy about what that might be, it’s worth noting. Sherlund also boosted his price target for Microsoft to $38 from $32 per share, while retaining his neutral rating on the stock. Hmmm.

This is fascinating because Sherlund, in his past life as software analyst at Goldman Sachs was “the” go-to analyst on Microsoft and he definitely knows a where some bodies are buried.

Here are some veiled semi-, sort-of predictions Sherlund put down in a research note released very early Tuesday morning:

1: Restive shareholders gain power: Sherlund thinks that shareholders are gaining steam in their demand for a greater voice in where the company is headed. There “may be a more receptive group of frustrated shareholders to leverage in an effort to drive greater realization of shareholder value at Microsoft,” he wrote.

2: Microsoft could exit search. It could hand search off to Facebook or Yahoo in return for traffic acquisition costs (TAC), Sherlund wrote. Microsoft Bing has gained some ground on Google but remains a distant second. The latest Comscore numbers showed Bing with a record 17.1 percent of U.S. searches in April, up from 16.9 percent in March. Google share fell to 66.5 percent from 67.1 perent percent for the same period. Whether that gain is worth what Microsoft poured into its no-doubt-pricey BingitOn campaign, is subject for another debate, however.

3: It could pay off disgruntled investors. Microsoft could double the dividend to yield about 6 percent by providing tax on currnt foreign source income.”

One thing is clear: Microsoft shareholders are one unhappy bunch. Looking at the post bubble-burst 13 years, the stock price is basically flat — it’s peaked at around $37 and has bounced beweeen that an d$20 for much of that time.

That’s led to some very loud calls for CEO Steve Ballmer to head for the door — something Ballmer shows no intention of doing. Dow Jones Newswire’s Al Smith helpfully published Ballmer’s Epitaph earlier this month, citing Windows 8 as a “bet the farm” gamble that didn’t pay off. That a contention — that Windows 8 is a failure — has been repeated in several news outlets. And it’s a conclusion that Microsoft’s top corp comms guy couldn’t let pass. In a blog post, Frank Shaw responded that Windows 8, which has sold 100 million copies, is hardly a failure. And linked to two positive reviews.

But back to Sherlund. He senses something different in the air when it comes to Microsoft’s corporate governance:

“We think there is a shift in the wind upcoming for Microsoft,with shareholders likely demanding a greater say in the direction of the company and how it might be run to drive a better return to shareholders.”

Now we’ll just have to wait and see.

MSFT data by YCharts

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.
May 28, 2013
Chinese compromise of U.S. weapon designs drives home painful lesson in cybersecurity
For anyone paying attention, the fact that Chinese hackers apparently accessed key U.S. weapons designs may be unsettling but hardly surprising. Previously undisclosed findings by the Defense Science Board show that more than two dozen major weapons designs were breached, according to a Washington Post report on Tuesday. Affected projects range from U.S. missile defenses to combat aircraft — including the F-35 Joint Strike Fighter — and ships. (The Post compiled a list of the affected weapons here.)

Dan Geer, a superstar among computer security and risk management experts, spoke to me about just this sort of risk last week. The most sobering part of the conversation was Geer’s stated belief that the game has definitively shifted from prevention of attacks to mitigation of their consequences.

In short: if you have something worth accessing, it will be accessed. The only realistic goal now is to make sure you know when that breach happens as fast as possible. I quoted him on this topic earlier, but his words ring even more eerily true now:

“If your enemy really is the People’s Liberation Army, what can you do? We can sputter about it but they’re serious and they’re good … The most serious attackers will probably get in no matter what you do. At this point, the design principal, if you’re a security person working inside a firm, is not no failures, but no silent failures.”

Of course security vendors have latched onto these threats as a way to sell more stuff and are increasingly glomming onto big data analysis as a way to shorten the time between an attack and stopping it in a high-stakes game of whack-a-mole.

As RSA executive chairman Art Coviello said a few months ago: “It’s not about perfect security; its all about ratcheting down risk as much as you can.”

And it’s not just huge government contractors, agencies and suppliers at risk. “No industry is immune,” cautioned Geer, who is also an advisor to In-Q-Tel, the investment arm of the CIA and other security agencies, and to Verdasys, a security vendor. Almost anyone can see why hackers target gigantic players like Boeing that spend billions on designs which could be used to build similar products at much lower cost. But don’t forget that any grocery store chain that uses credit cards is also a target for someone, Geer said.

Feature photo courtesy of Flickr user Dysanovic

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.
May 28, 2013
The week in cloud: FUD and loathing edition
Fear, Uncertainty and Doubt (FUD) has been a factor in technology adoption since, well since forever. And that there’s FUD around cloud adoption driven by data security concerns is a no brainer. Sure, companies can go for all sorts of security certifications, but it’s unclear that accreditation will ease cloud anxiety.

Fanning the FUD flame this week was a new report (PDF here) from the Commission on the Theft of American Intellectual Property estimates that IP theft costs the U.S. economy $300 billion per year. Not surprisingly, China was fingered as the chief culprit, although Russia and India were also named as problem children here.

The panel, Reuters reported, wants to anoint the president’s national security advisor as grand poobah of a ramped-up effort to protect intellectual property. Bad behavior should be met with banking sanctions, import bans and financial blacklisting, according to the 89-page report. (In related news, former CIA Director James Woolsey this week warned that the U.S. is at risk of a cyber attack from North Korea which truly is scary. At least Chinese motives seem rational.)

Security, or lack thereof – is one big de-motivating factor when it comes to moving more corporate workloads to the cloud– or allowing employees to use their personal smartphones at work for that matter. But it would foolish for security vendors to waste a good crisis.

This week Verdasys, for example, launched an update to its managed security service to extend protection to users’ end-point devices. Expect a raft of similar offerings and updates to come.

Security guru Dan Geer says high-stakes hacking is now a way of life and the best any company can do is to mitigate, not eliminate, risk.

“If your enemy really is the People’s Liberation Army, what can you do? We can sputter about it but they’re serious and they’re good,” he told me in an interview last week. “The most serious attackers will probably get in no matter what you do. At this point, the design principal, if you’re a security person working inside a firm, is not failures, but no silent failures.”

The key is to know as soon as possible that a breach has occurred and to react fast.

Other than that, key design points are to keep things patched and to put firewall filtering both on incoming and outgoing traffic, said Geer.

Other key cloud news from GigaOM and elsewhere

It was a very busy week in cloud land. Here’s a recap of GigaOM’s coverage:
From PC World: Nvidia, Citrix crank up virtual desktop delivery

From The Guardian: Public,private or hybrid cloud.

Citrix Synergy 2013 live blog from Brian Madden.

And now, for your moment of Zen

With apologies to Jon Stewart:

Photo courtesy of Flickr user annabellaphoto

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.
May 26, 2013
VMware’s revolving door keeps on spinning
A week after VMware announced its top-priority Infrastructure-as-a-Service play, two more executives associated with de-emphasized technologies have left the company.

VMware CEO Pat Gelsinger

Both execs, Javier Soltero and Kevin Henrickson joined VC company Redpoint Ventures as entrepreneurs in residence.

Soltero joined VMware by virtue of its acquisition of SpringSource in 2009 — after having joined SpringSource by virtue of that company’s acquisition of Hyperic. At VMware he was the CTO of SaaS and application services.

Henrickson was senior director of R&D for Zimbra, the open-source email product VMware acquired from Yahoo also in 2009. The dual departures were first reported by TechCrunch.

It was clear last year that VMware was scaling back on applications – which had been a key part of former CEO Paul Maritz’s strategy. It subsequently spun off the Java-based Spring framework along with Cloud Foundy and other assets to Pivotal where they will be part of that company’s universal PaaS push.

Selling off the non-essentials

It’s fairly clear that VMware would like to divest itself of Zimbra, which doesn’t fit into its new IaaS worldview, just as it sold off SlideRocket to Clearslide and Wavemaker Java technology to Pramati, both in March. (Pramati just announced Cloudjee a new company pushing a cloud development platform incorporating Wavemaker technology.)

Early this month, Patrick Chanezon, who led developer relationships for both Spring and Cloud Foundry efforts at VMware, joined Microsoft as director of enterprise evangelism.

We can’t say that VMware didn’t warn us. In January, CEO Pat Gelsinger clearly stated the company’s need to focus and eliminate distractions. At about that same time CTO Stephen Herrod left the company for General Catalyst.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.
May 24, 2013
How Amazon’s cloud competitors are trying to find cracks in AWS’s armor
It’s not exactly shocking that Amazon cloud competitors are polishing up their PR talking points about the benefits of hybrid cloud. And turning up the volume on their pitches.

Here’s why: As Amazon Web Services keeps churning out services, support offerings and certifications to appeal to corporate and government users (the latest being FedRAMP accreditation), other cloud vendors need to show that they offer value above and beyond AWS. Hybrid cloud, which pairs local processing power with outside cloud resources as needed, is one area that they see as a weakness for Amazon.

AWS versus everyone else

While none of these rivals refer to themselves as AWS killers (smart move), they all see Amazon as the #1 cloud player and the top threat to their own cloud ambitions. When pressed, VMwarez senior vice president Matthew Lodge acknowledged that “everyone is competing for the same IaaS dollars.” Everyone meaning Amazon and the rest of the cloud contenders.

VMware, which saw, um, limited uptake of the vCloud Director that it pushed service providers to use as the basis for their own clouds, said its new vCloud Hybrid Cloud Services will compete with AWS on price, at least in some cases, but offer other enterprise-worthy goodies.

Said Lodge: when you factor in “hidden costs” in Amazon’s dedicated instances, the playing field levels out. “They charge for I/O and we don’t. They charge for VPN endpoints, load balancers and firewalls and we don’t,” he said.

Rackspace president Lew Moorman has a similar message. “Now that public cloud is 3 to 4 years old in reality, applications are bigger and more complex and people are starting to see tradeoffs to using public cloud only,” Moorman told me Thursday.

Structure 2012: Lew Moorman – IT Cloud Lead, Intel Corporation

“When public cloud came out and you could suddenly provision a server in a minute when it used to take 3 months, those were intoxicating advances … you get drunk on them but when things settle in there are tradeoffs,” he said.

For examle, what’s great for test-and-dev environments is not always optimal for production workloads, where public cloud costs quickly add up.

Once someone hits the $25,000-a-month milestone, “it’s time to rethink all-public-cloud deployment,” he said.

Joyent trumpeted a similar message this week when it announced a raft of new compute instances it says will be competitive with AWS. Joyent, like Rackspace, offers public, private and hybrid cloud options.

Corporate cloud purchases are about more than price and technology

Having said all that, almost every cloud vendor alive will also add that price isn’t the compelling reason to move to cloud. Face it: when it comes to IT-sanctioned technology purchases, it’s not just about the price or the technology. IT departments have established procedures and guidelines for deployment and cloud providers will have to accommodate them.

“Most public clouds — AWS etc. — don’t offer enterprise-class security, compliance or performance SLAs to users,” said Rodney Rogers, CEO of Virtustream, which positions itself as an enterprise cloud provider. ”Some public clouds offer supplemental services that dedicate equipment to enterprises/government, but they are generally not multi-tenant and so deliver less efficiency.”

That means they remain suited for test and dev, for backup, SaaS apps and apps with no performance criteria, Rogers said via email.

Is Amazon’s head start insurmountable?

Granted all of this is self serving talk, but having sat through a raft of CIO panels this week, it is clear to me that some of these points ring true with this constituency. But, if we’ve learned anything from the past 6 years of its existence, AWS won’t stand still. It now offers several services of its own and through an alliance with Eucalyptus that break down some barriers between a customer data center and its cloud. But until you can run AWS instances on your own infrastructure, AWS will remain a public cloud provider in a world where more workloads could flow to a hybrid model.

Structure 2010: Werner Vogels – CTO and Vice President, Amazon

AWS has a huge head start and lots of customers. But we’re early in the cloud era. IDC says less than 5 percent of the world’s total IT budget is now devoted to public or private cloud. That leaves a lot of upside for Amazon and its competitors.

There’s time for Amazon to offer more hybrid options and for rivals to catch up. It’ll nothing if not an interesting market over the next few years.

Who wants to bet that this topic of hybrid vs. public cloud deployment will come up at Structure 2013 next month where both Moorman and Amazon CTO Werner Vogels will take the stage?

Pretty safe money, I’d say.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.
May 24, 2013
Tippy top stars of Techstars Demo Day (Boston Edition)
Techstars Boston Demo Day was in a glitzy new setting (the House of Blues withing spitting distance of Fenway) and also drew some surprising (non-tech) star power.

Here are my highly subjective highlights:

1: David Ortiz.

A (very fuzzy) David Ortiz at TechStars Demo Day.

There was a bona fide Big Papi moment on stage as the Red Sox superstar and Boston superhero, in general, as David Ortiz strode on stage to greet Fancred CEO Kash Rassaghi. This startup is building a “social platform” to connect sports aficianados with like-minded fans. Ortiz demanded that Rassaghi “get the Yankess off my cell phone.” (I dropped my phone but recovered in time to get one sub-par shot at left.)

Is there really room for a sports fan platform? Doubtful. But, hey, I’ve been wrong before. And did I mention DAVID ORTIZ???

2: A platform for sustainable, local food

I love the idea behind Freight Farms, which takes shipping crates and retrofits them with water, electricity, internet access and LED lighting to convert them into compact hydroponic gardens.

The elegant idea is to “take the very structure that makes the global food supply chain possible and make it into a platform for producing local food,” said Brad McNamara, Freight Farms CEO.

Freight Farms CEO Brad McNamara.

They are remotely controlled and, because they are stackable, they take up less real estate. Freight Farms has signed several customers including Katsiroubas Brothers, a 100-year old Boston-based produce wholesaler, which is looking for better ways to cut transport costs and offer customers more local product.

“There is nothing better than fresh local food, but the reality is food distribution is a long complicated supply chain — most goods travel 1,500 miles on average to get to your table,” said McNamara.

Freight Farm-grown crops require less water, no pesticides or herbicides. My question: Will a their tomatoes taste like other hot-house tomatoes (i.e., like cardboard) or like an actual tomato? If it’s the latter, I’m totally sold.

3: DIY clothing design

In a nod to the burgeoning “maker market” or do-it-yourself crowd, Mary Huang was to hand to talk up Constrvct, her startup that’s building service that lets you design 3-D clothing onscreen, tweak the size and styling with easy slidebar controls, preview your design on an onscreen maniquin and then make your clothes to order.

“Makers are underserved in the do-it-yourself market — they’re stuck at the same starting point as their grandmothers,” Huang said. Interest in home-designed clothes is rising thanks to Pinterest and Project Runway, she said, quoting a surprising stat: 3 million sewing machines sold last year, double the number from ten years ago.

4: Fixing manufacturing

LinkCycle says it can use its own data science — and existing data — to help manufacturing plants drastically cut their energy costs.

These facilities — many of them rust belt relics — are notorious for wasting energy and to remedy that many spend millions installing meters and hiring auditors to help. Most of that spending is also a waste, according to LinkCycle CEO Sahil Sahni.

“Why spend so much time gathering data when companies are already sitting on heaps of it?” he asked.

LinkCycle instead takes two exisitng data streams from the ERP systems already running these companies — electricity consumption and total production output. “We developed our own algorithms to take that data and use math — not meters — to save money wihtout having to set foot in the plant,” he said.

Wow, that sounds so easy it makes you wonder why someone else hasn’t done it. Well except for that algorithm part anyway.

So, the new venue was fab but it suffered the same woe as past Techstars events — a lack of reliable connectivity. We soldiered through with personal hotspots and (finally) some intermittent Wifi connections but can’t one of these deep-pocketed sponsors finally figure out how to get reliable broadband into these events? (I’m looking at you Microsoft, Rackspace, Verizon and Softlayer.)

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.
May 23, 2013
Google cuts prices on week-old datastore
Let the cloud price chopping continue.

A week after Google announced its Cloud Datastore managed NoSQL database, it’s cutting its price, according to a Google blog post.

That means a 25 percent lower prices in some cases for both CloudStore per se and Google App Engine’s High Replication Datastore (HRD).

So that’s the latest cloud price cut, but it won’t be the last. Next up: Amazon? Microsoft?

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.
May 23, 2013
Joyent to Amazon: It’s on
Okay, it’s a bit of a David and Goliath story – Joyent is a cloud provider that seems to maneuver just below the radar. But on Thursday it will come out fighting with an array of new compute instances — including reserved instance pricing — to position itself as an attractive alternative to big, bad Amazon Web Services.

Joyent CEO Henry Wasik

San Francisco-based Joyent has made noises about going up against Amazon before but now it’s more than tripled the number of instance types it will offer, including 7 different “standard” instance types with RAM allocations ranging from 0.5 to 128 GB; 5 high-memory instances; 6 high-CPU instances; 3 high-storage instances; and 3 high I/O instances (see chart.) But that’s just the beginning, said Joyent CEO Henry Wasik, who joined the company in November.

“We’ve completely reformatted what we do and dramatically expanded the number of instances — originally we had 10 and now 27, but once the portal is turned we’ll have 71,” said he said.

Depending on the workload, Joyent services may well be cheaper than AWS, he said. (Stay tuned for Amazon’s response.) But as many have pointed out, for cloud providers, competing on price alone is a fool’s errand.

Joyent seeks to differentiate itself on how well it runs high-performance applications on its own SmartOS (or on Linux or Windows); the tooling it provides; its service and support; and its ability to offer the hybrid cloud option that many companies prefer.

Earlier this week Dell said it would offer Joyent as one of three public cloud options it will sell to customers. Dell had promised to deliver an OpenStack-based public cloud this year, but thought better of it.

Face it, in the cloud computing world, it’s Amazon first and then everyone else. In one of my favorite posts of the year comparing cloud providers to hamburger franchises, GigaOM’s Derrick Harris posited that AWS is McDonalds, Rackspace is Wendy’s but a handful of providers — Joyent, Virtustream, CloudSigma — represent the In-N-Out Burger (yum!) or Five Guys of cloud. He wrote:

These cloud providers, like their analogous restaurant chains, are damn good at what they do and their patrons are loyal. They’re typically designed for maximum performance, maybe security, too, and will play around with new infrastructural or programming components in order to maintain their edge. They might even be the best at certain things and have some major customers (I’ve seen Maseratis leaving the In-N-Out drive-thru), but cost, geography or the desire to get a chicken sandwich, too, limit the number of users they can attract.

I know that we’re early on in cloud adoption and that the potential workloads moving to cloud is high. But to me it’s clear there will be a shakeout as enterprise players like VMware — which announced its public cloud option this week — along with Dell, IBM, HP and Red Hat try to preserve their traditional IT strengths in a cloud venue while newer look players built for the cloud — Joyent, Virtustream, and others — gear up.

There may be a ton of work out there but i would bet that some of these players will not be standing in two years’ time.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.
May 23, 2013
6 things every CIO should know (or at least think about)
CIOs have a tough gig. They’re besieged by the bring-your-own-device (BYOD) boom; have to keep up with the latest cloud services; they have to assess the value of the latest big data innovations; and often deal with CEOs and subordinates who think they could do the job better.

“Being the CIO of a tech company is easy, you support 60,000 users and 59,999 think they have your job,” EMC executive vice president and COO (and former CIO) Sanjay Mirchandani told attendees of the MIT Sloan CIO Symposium on Wednesday.

Here are the key takeaways from the conference.

1: Learn to work well with others

Erik Brynjolfsson, director for Digital Business at MIT Sloan School of Management .

It’s a lesson we were all supposed to learn in grade school but too often IT staff and the business units beyond are not fully in concert. That’s got to stop.

CIOs are tired of being “the department of naysayers” and the best ones embed themselves in the business cases of the company. And they have to be active participants in business talks, not order takers for technology.

In the traditional model with CIOs, it was “come in and tell me what your requirements are — we need to get out of that we need to have the courage to participate in the process,” said Michael Loo, SVP of Global IT for Avaya.

2: Embrace big data but don’t be bedazzled.

Everyone is besotted by data. The more the better. A common mistake is to confuse correlation with causation, was another refrain. But without the right background knowledge and statistical analysis tools you can still leap to the wrong conclusions about your data.

“There’s a shocking correlation between lung cancer and people who have ashtrays at home,” said Andrew Lo, professor of Finance at MIT’s Sloane School of Management.

Alex “Sandy” Pentland, professor at the MIT Media Lab concurred. “Big data is good for interpolation, when you know the field you’re working in but bad for extrapolation where you’re entering new areas,” Pentland said.

3: Stop managing by gut

Having said that, there is still no substitute for data and metrics, properly applied and analysed. Too often early in the war on cancer, medical research has been hobbled by researchers’ tendency to let intuition drive their trials, said Dmitri Bertsimas, professor of operations research and statistics at MIT’s Sloan School. “We didn’t make a lot of progress,” he said.

“We need to change from opinions and hunches and go with facts and data,” agreed Erik Rynjolfsson, director of the MIT for Digital Business at the Sloan School.

4: Beware the HIPPO

Too often key business and tech strategic decisions are made by the HIPPO, aka the ”Highest-paid Person with an Opinion.” That person may big-foot a process where discussion and pushback are a better to go.

This is a term I first heard a few weeks ago at another industry event where Phil Swisher, chief innovation officer at Brown Brothers Harriman bemoaned this tendency of HIPPOS to dominate discussion. As The Boston Globe’s Scott Hirsner wrote:

Running experiments is much better than simply taking direction from a HIPPO, as politically difficult as that may be. “Hypothesis testing is better than hunches,” Swisher said.

5: Balance innovation with stability.

Remember, bleeding edge is bleeding for a reason. Sometimes you do have to simply make sure the trains run on time, while hopefully making those trains better, faster, cheaper over time.

Kazuhiro Gomi, president and CEO of NTT America, knows from experience. “Many of our customers rely on us to run their systems for them and running things smoothly is sometimes more important than being innovative,” he said.

6: Beware the CMO

There’s been a lot of talk that chief marketing officers will control more of the IT spend than CIOs. “Over my dead body,” said Avaya’s Loo said he was well aware of the study which also said “CMOs and CIOs should be best friends.” That makes sense, Loo added, because CMOs need data and to get data they need things like point-of-sale systems and e-commerce sites which are all about IT.

So, will CMOs get more IT spend then CIOs? Loo said: “Over my dead body.”

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.
May 22, 2013
Need a quick $30K? Recommend the right developer and it’s yours
How much is a truly great developer worth? A lot, according to Hubspot and it’s willing to pay a $30,000 bounty to anyone (anyone!) who refers a great developer that the marketing automation software maker ends up hiring.

The company’s pitch:

HubSpot loves developers. Like, heart-pounding, chest-thumping, breathless love. Care to play matchmaker?

We’ll give you $30,000 if you refer us to a developer who makes us feel that way.

The company has openings in its Cambridge and Dublin locations and has cash to spend. It netted $35 million in a financing round last November, bringing total funding to about $100 million. One of the priorities for that money was staffing up, CEO Brian Halligan said at the time.

We know that great developers and engineers can write their own tickets, but $30,000 is a big number to give to a friend of a developer. It’s also a great way for any software developer to build a database of potential hires down the road. I’ve heard of smaller incentives before – Hubspot itself once offered $10,000 — but this is a lotta scratch.

If you know of others please let me know in the comments section below.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.
May 22, 2013
OpenStack cloud builder Cloudscaling nets $10M from Juniper, Seagate and friends
Cloudscaling, a company that’s made its name actually building OpenStack clouds for real customers including Korea Telecom, just closed a $10 million Series B round including money from new backers Juniper Networks’ Junos Innovation Fund and Seagate. Trinity Ventures, an early investor, also participated.

The San Francisco-based company targets new applications that run on its technology but can also run in Amazon and Google public clouds as needed. Juniper is already partnering with Cloudscaling which is using Juniper’s Virtual Network Control in a Virtual Private Cloud (VPC) feature that maps to virtual private clouds in Amazon Web Services.

The funding, “affirms that customers want more than OpenStack. They want an on-premise, OpenStack-based private or public cloud turnkey system solution that delivers architectural and behavioral fidelity with major public clouds like Amazon Web Services,” Cloudscaling CEO Michael Grant said in a statement.ing

It’s been a sort of rocky week for OpenStack. On Monday, Dell surprised the ecosystem by deep-sixing plans to roll out an OpenStack-based public cloud.

Cloudscaling closed a $4 million Series A round in September 2011. Company CTO Randy Bias, pictured above right, is an OpenStack Foundation board member, and will talk about issues and benefits of cloud adoption at GigaOM’s Structure Europe in September.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.
May 22, 2013
Cloudcheckr boosts support for Amazon GovCloud
CloudCheckr, one of several vendors that monitor Amazon Web Services usage for customers, says it is the only one of those rivals that can do that job for Amazon’s restricted GovCloud. GovCloud is a separate U.S. region set up for state, local and federal agencies that must meet special requirements for cloud use.

Tools like CloudCheckr’s service can help in the government procurement process — a big deal given the U.S. government’s cloud-first mandate, which requires agencies not only to deploy a different sort of technology, but to readjust how they think about buying and paying for services.

“They have a hard time dealing with cloud costs because they’re so used to fixed-cost contracts,” said James Hirmas, COO of JHC Technology, an AWS consultancy specializing in government work and a prime contractor for the National Institute of Standards and Technology (NIST). JHC worked with CloudCheckr to integrate its service with GovCloud.

With that integration, a customer can see if it’s underutilizing compute instances for a certain task and, if so, advise that the work be moved to a smaller, cheaper instance, for example. CloudCheckr performs compliance checks and best practice analysis for GovCloud environments.

Aaron Klein, COO of Rochester, N.Y.-based CloudCheckr, said the GovCloud service does 90 percent of what it does on the commercial side. “GovCloud is architected differently from other AWS regions,” he said. “First you need access, then you need to delve in and adapt what you have to work best in that environment.” He also pointed out that not all of AWS’s own services are running on GovCloud so far.

Since GovCloud is compliant with the International Traffic in Arms Regulations (ITAR) – only U.S.-born personnel can work there or access it. Its help desk is U.S.-only. Background checks are also required.

Amazon itself is clearly gearing up for more government work, having received its FedRAMP certification early this week. That accreditation should make it easier for more government entities to use GovCloud (or other U.S. regions depending on the workload) without having to go through a lot of redundant testing and paperwork.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.
May 22, 2013