Another little IBM deal
February 11, 2009
On August 12, 1981, 28 long years ago, IBM introduced its personal computer, the IBM PC. Hidden inside was an operating system called MS-DOS which the computing giant had licensed from a pipsqueak company named Microsoft. IBM didn't realize it at the time, but the deal, which allowed Microsoft to maintain its ownership of the operating system and to license it to other companies, turned out to be the seminal event in defining the commercial landscape for the computing business throughout the ensuing PC era. IBM, through the deal, anointed Microsoft as the dominant company of that era.
Today, as a new era in computing dawns, IBM announced another deal, this time with Amazon Web Services, a pipsqueak in the IT business but an early leader in cloud computing. Under the deal, corporations and software developers will be able to run IBM's commercial software in Amazon's cloud. As the Register's Timothy Prickett Morgan reports, "IBM announced that it would be deploying a big piece of its database and middleware software stack on Amazon's Elastic Compute Cloud (EC2) service. The software that IBM is moving out to EC2 includes the company's DB2 and Informix Dynamic Server relational databases, its WebSphere Portal and sMash mashup tools, and its Lotus Web Content Management program ... The interesting twist on the Amazon-IBM deal is that Big Blue is going to let companies that have already bought software licenses run that software out on the EC2 cloud, once the offering is generally available."
Prickett Morgan also notes, "If compute clouds want to succeed as businesses instead of toys, they have to run the same commercial software that IT departments deploy internally on their own servers. Which is why [the] deal struck between IBM and Amazon's Web Services subsidiary is important, perhaps more so for Amazon than for Big Blue."
It doesn't seem like such a big deal, and it probably isn't. But you never know. The licensing of MS-DOS seemed like small potatoes when it happened. Could the accidental kingmaker have struck again?
UPDATE: Dana Gardner speculates on the upshot.
Though IBM's unbundling decision in 1969 gave birth to the Independent software industry, I dont think they enjoy the same kind of power or influence today. There are plenty of providers of the cloud and Amazon is definitely the first choice. But that's about it. Though I always wonder, with all the hardware that IBM manufactures and massive computation efforts like the Bluegene, why isn't IBM building their own cloud?
Posted by: me.yahoo.com/riteshmnayak at February 12, 2009 04:14 AM
I agree that the situation is very different today than it was in 81 (or 69), but IBM still carries both real and symbolic influence over corporations and its imprimatur will help buttress the credibility of the cloud model. That may turn out to be a modest force in the adoption of the cloud; on the other hand, you never know what event will push a trend over the fabled tipping point.
"why isn't IBM building their own cloud?"
It is, but I think they see the opportunity in terms of outsourcing and consulting - big-margin services. You have to remember that Amazon's model is low-margin, which makes it a difficult one for big IT vendors to swallow.
Posted by: Nick Carr at February 12, 2009 09:36 AM
"why isn't IBM building their own cloud?"
The same reason they didn't build their own operating system back in '81... their corporate culture still, despite many years of reform and adaptation, doesn't breed teams that can put together these sorts of systems in reasonable timeframes.
Until recently, Amazon was still calling most of their web services "beta" products. Can you imagine IBM throwing together something on that scale and releasing it as a beta? But that's what you have to do to grow systems of this sort.
As Nick says, IBM is indeed working on their own cloud, but you would barely recognize it as such compared to everyone else who is putting together utility computing services... it's cramped by the same tradition and complication that kept the company from developing a working micro-computer operating system inside the tight timelines the PC project required. The PC was an aberration for the company, but it was the beginning of an era that required more of the same: faster, less tightly bound development cycles and releases. Utility computing is only accelerating that trend and IBM is still behind the curve.
Posted by: Scott Wilson at February 12, 2009 09:53 AM
Too little, too late on IBM's part. EC2 already has great mindshare if not marketshare. There is no fear of this particular tie up being a huge success. DB2, WebSphere Portal and Lotus are not exactly the trademarks that excite battle hardened application developers and systems integrators of today.
Posted by: Sriram Narayan at February 12, 2009 01:35 PM
I enjoy reading Nick's commentary. He's also kind of cute to boot.
You are right that this is significant, imo.
The particular product here - the deal - the scheme - might stand or fall. My guess is "fall" but only barely. The key thing is that IBM and Amazon or working out the right biz-model.
"Commodity" is "commodity" and apps are a separate thing. IBM is just pro-active here in porting some of their more interesting "legacy" products to the new cloud platform in advance of the "cloud" being particularly well defined.
Reading recommendation: there's something today (the day the subject article was posted) on Radar (radar.oreilly.com) that links work going on in the RAD lab at CAL (that's the illustrious lab associated with Patterson at U.C. Berkeley). I've only skimmed it so far but it looks very promising (as one would wish for / expect from a Patterson-related production). (anti-disclaimer: no, i've no connection or interest in the RAD lab or Patterson).
What I think we're going to wind up with, as an overlay over the internet but where most of the action is, is a "service-addressed-network" as opposed to the currently dominant and never-quite going-away "host-addressed-network". When you "google" something that won't mean paying tribute to "google.com" but, rather, will mean leasing some tiny amount of time on a semi-generic cloud server running the "google.exe" program, so to speak.
At first glance, it looks like Patterson et. al are thinking in clever ways about how such essential infrastructure gets monetized (but again, I've only skimmed - can't vouch for the meat of the Radar cite.)
Posted by: Tom Lord at February 12, 2009 07:19 PM
Post a comment
Thanks for signing in, . Now you can comment. (sign out)(If you haven't left a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Thanks for waiting.)
"Riveting" -San Francisco Chronicle
"Rewarding" -Financial Times
"Ominously prescient" -Kirkus Reviews
"Riveting stuff" -New York Post