Dean Michael Berris

Earlier today I had been reading from Twitter that IBM and Sun were in acquisition talks. Personally, I wouldn't really care because neither am I a businessman (yet) nor a customer of either companies -- yes, I don't develop in Java -- but I do have a few things to say about whether I think it's a good thing if it happens and what are the possible consequences of such a merger. Also, I agree with Om Malik that Cisco should be the one talking with Sun regarding an acquisition for more than just a handful of reasons. In this post I will detail a few of these thoughts along with my personal take on some of the articles I've read about the rumored talks.

The Cloud

Anybody who's paying attention to technology nowadays has pretty much heard something about cloud computing and how it's changing the way the web and enterprises think about computing. For instance, Amazon has been trying to get enterprises lately to move their infrastructure into Amazon Web Services. Another player that's triyng to leverage on cloud computing is Google which has shown a focus on web-based applications that should be able to run well in the Google infrastructure with the Google App Engine.

What's not immediately clear about this whole cloud computing initiative is that inside the whole cloud thing are computers. And the computing power in the cloud doesn't just pop out of nowhere -- it comes from the computer manufacturers and the server specialists. In this level the manufacturers are jockeying for position to secure contracts to supply hardware at massive scales (hundreds and thousands of new machine sales are on the line). There's support contracts to be signed, new hardware to be made, new software to be written to handle these new machines, and the whole infrastructure conveniently hidden behind the cloud has got to be powered by computers made by someone at some point.

Google as far as I know doesn't buy pre-made computers, but rather they build the computers they use from components on their own -- so pretty much they're not a customer of Sun or IBM. Microsoft on the other hand might be a customer of Dell or HP but I'm not sure if Sun or IBM actually supply them with their IT requirements. So where am I getting at with this?

I.B.M. is in talks to buy Sun Microsystems in a proposed deal valued at nearly $7 billion, a person with knowledge of the negotiations said on Wednesday.

The merger, if completed, would be a major consolidation in the market for server computers used in corporate data centers — and one that could prompt an antitrust challenge.

For I.B.M., the move would be something of a departure from its successful strategy of paring its dependence on the hardware business, where profit margins have declined, and increasing its investment in higher-margin software and services businesses.

But the combination of I.B.M. and Sun, analysts say, would bring together two technology companies that have continued to invest heavily in research and development, when many of their corporate peers have cut back to shield profits. I.B.M.’s research and development budget is about $6 billion a year, while Sun’s is roughly $3 billion.

I.B.M. Said to Be in Talks to Buy Sun for $7 Billion
New York Times 

For all the other companies who aren't Google, Amazon, or Microsoft, they've got to buy computers, software, and hire consultants and get services from either IBM or Sun. They serve the same market pretty much and gear their solutions towards the enterprise whose businesses aren't to offer technology but rather use technology to get their work done.

Think of banks, retail stores, hospitals, stock markets, governments, and pretty much any industry where you will need computers to store, manage, and deal with corporate or internal data and you'll find either one of IBM or Sun. It's no surprise that people are talking antitrust early in the game because when once you had a choice among Sun, IBM, HP, and other (smaller?) players, you now had to compete with IBM+Sun who would combine for a whopping 65% of the enterprise computing market.

So that cloud the enterprise might be building might be Big and Blue and is hiding a Sun.

The Technology

There are a whole slew of technology that's hidden within IBM and Sun that's been relevant in the new world influencing the very lives we lead. Let me just list a few of these technologies that I can think of off the top of my head:

• Java - the first programming language that made the Internet a place where you might be able to do business in. Before Java, there wasn't a very big Internet agenda in the corporate scene. And before Java too, the notion of a "dynamic web" wasn't very alive. Although it's not the prime player today (although it's still strong) Java can now be found in almost every device you have which allows you to run rich applications on these devices.
• PlayStation - the console that revolutionized the whole console gaming industry. After the Nintendo game consoles stagnated a bit, Sony jumped in and filled that gap. The latest incarnation of the playstation is now powered by IBM-developed processors which represent the next generation of high performance computing at a low price point. If you have a PS3 in your house right now, you can bet you have a supercomputer sitting idly by occasionally allowing you to enjoy your games and Blu Ray movies.
• The PC - of course the whole PC industry was started by IBM. No need to expound much about that but think about the way it changed how we do things in our daily lives and how it changed the way business is done worldwide. Without the PC the whole "information age" wouldn't have happened.

I'm sure others can list a lot more things that IBM and Sun can be proud of but what else could happen when Big Blue assimilates the Sun?

The two companies have a common interest in that both make computer systems for corporate customers that aren't reliant on Microsoft Corp.'s Windows software or Intel Corp.'s microprocessor technologies. The two companies are also strong supporters of open-source Linux and Java software for Web application development.

IBM in Talks to Buy Sun in Bid to Add To Web Heft
Wall Street Journal

That's right, a bigger push for Linux and Open Source Software in the enterprise world. It's been happening for a while but I'd think a bigger Big Blue can push its Linux and Java agenda to more customers of Sun. While Sun has the OpenSolaris platform that works well on Sun machines that also contain the Intel chips, the choice for the hard-core Unix users would be easy. And when it's backed by the IBM consultants and partners, you've got a formidable contender for "the single IT resource you'll ever need for your data center and application development needs".

The Problem

So if you look at it from this vantage point, it seems to be a good thing for IBM to buy Sun and for Sun to be bought by IBM. After all, they're pretty much compatible when it comes to the market (consolidating the customer base that use pretty much the same solution stack anyway). But then there are a few problems.

• Sun is not making much money at this time. IBM may be buying a company that will have trouble sustaining itself if it goes on the same path it's going at the moment. Moving into the cloud is good, but making money off that strategy may take a while.
• The antitrust problem is a big one. IBM has already gone through one breakup before and going into another one as a result of this acquisition spells disaster for the business too.
• Slash and burn. IBM may be inclined to just buy Sun and keep the parts it likes and sell of the other parts it doesn't like. This may effectively mean that if it does happen, you can kiss Sun (the brand and company) goodbye.

I do however tend to agree that Cisco should buy Sun instead. Quoting Om's take on it:

I firmly believe Sun is a great engineering resource. The problem would be running the place, as anyone who bought it would be taking on a serious management challenge. Still, I can name a few potential suitors — Hewlett Packard, Dell, Juniper Networks or Hitachi, or even private equity guys (led by former Sun co-founders) — that would be up to the challenge. Damn, if I had the money, even I would try to buy it.

But no one is better equipped to buy and manage Sun than Cisco Systems. Despite its bold entry into “unified computing” with the launch of a new line of blade servers, Cisco is a newbie in the field and its product line so far is limited. Sun has a vast array of products, not to mention significant software resources (MySQL) for Cisco’s vision of unified and cloud computing and the sales channels and existing customers to help execute it. (IBM has all those things already, and it can pick up new customers via Sun.) More importantly, it has engineers — the very people Cisco values most and historically a strong factor in the networking giant’s acquisition decisions.

Why Cisco, Not IBM, Should Buy Sun
GigaOm Network

Another good argument that Erick Schonfeld makes for the IBM and Sun deal is that:

The acquisition would create an even greater mish-mash of technologies under the IBM umbrella, but IBM has never had a problem selling incompatible systems.  Sun also has a deep wealth of technology patents and engineering talent that IBM will be able to redeploy.

Big Blue Wants To Swallow Sun For $6.5 Billion
TechChrunch

So this looks like a good deal for IBM, but not such a good deal for Sun if they get acquired.

End Game

In the end I definitely think Cisco would gain more if they acquired Sun. Their wanting to enter the cloud computing arena and the recent Sun strategy to move into the cloud seems a compelling investment to make to solidify Cisco's prominence in the data center. However granted that IBM would like to have a better set of technologies it controls to offer to customers (current and potential), I think IBM will try hard to make sure they get Sun before anybody else does.

I personally would like to see the SPARC processor get a new lease on life in this new virtualized/parallel world either as a multi-processor beast or (if IBM can pull it off) as an off-board co-processor much like what the GPU's of today are. It shouldn't be hard to imagine a plug-and-play compute/storage cluster that grows with compute nodes in the price range of PS3 units (running some modified version of the Java virtual machine tuned for these new architectures).

If Cisco though had their way prepare to see stackable and plug-and-play do-it-all boxes for a homogenous cloud all running Linux on SPARC processors. Oh, and they're routers and NAS'es at the same time. Need more power? Just plug in a few more "Cisco Sun X Node" machines. Good times for the Bay Area if it happens.

Exciting times ahead -- can't wait to see what happens soon!