Why Is IBM, Not Cisco, Buying Sun?

Commenting on a pending acquisition is like touching the proverbial third rail, as with politics and religion. No one wants to be perceived as trying to influence the outcome of a deal that’s still in the works—especially when it’s unconfirmed—such as IBM’s reported $6.5 billion (£4.4 billion) bid for Sun Microsystems.

Since the news broke that IBM might be buying beleaguered Sun, there has been a steady stream of whispers that this is nothing more than any of the numerous rumours of big vendors buying other big vendors, insert corporate names here.

Adding validity to the rumour is that it comes in the wake of Cisco Systems’ entry into the data centre with virtualized blade servers and its unified computing architecture. Because Cisco is stepping into the backyards of its allies IBM, Hewlett-Packard and Dell, it only makes sense for one of the big three to make a major move to stave off competitive pressures. In this case, for IBM to make a play for Sun.

While many are questioning the wisdom and validity of the purported IBM acquisition, the bigger question to ask is this: “Why isn’t Cisco buying Sun?”

Cisco has made its intentions clear through its unified computing announcement that it wants in on the server market. And it’s no secret that Cisco’s road map will likely take it into storage next, and that would mean an expensive acquisition of either EMC or NetApp. So why not just bite down now and get deep into both servers and storage, as well as few other things, by picking up Sun?

On paper, the IBM-Sun deal looks like a grab for market share. Sun is the fourth-place supplier of servers, with roughly 10 percent market share. IBM is the market leader, with 31 to 33 percent. Buying Sun would put it well out of reach of second-place supplier HP, which holds roughly a 30 percent share of the server market.

But what else would IBM get for its $6.5 billion (£4.4 billion)? Storage? It already has that. Storage and systems management? Has that too. A database, since Sun owns MySQL? IBM already has DB2. Identity management and directories? Tivoli has plenty of that already. Processors? IBM already has the P5, which is akin to Sun’s SPARC.

Perhaps the only thing IBM would get that’s new is Java, the popular Web development platform. But is that worth the price of admission? Will that be a billion-dollar business?

Worse, an acquisition would foul several of IBM’s existing OEM partnerships. IBM is partnered with NetApp for NAS (network-attached storage), which would complicate things if longtime rival Sun were introduced. And Sun is partnered with Symantec for storage management and backup applications, which would lead to conflict with Tivoli.

Some analysts are beginning to speculate about the potential for antitrust scrutiny of an IBM-Sun union. Placing more than 40 percent of the server market in IBM’s hands is one thing, but analysts believe federal regulators—and some competitors—will oppose the acquisition since it will give IBM 65 percent of the Unix-based server market and leave only one significant competitor: HP.

If Cisco bought Sun, it would likely draw little, if any, regulatory scrutiny since any market-share capture by Cisco would be new and would not result in a net loss of competitive players.

Cisco could do well to pick up Sun and jump into several markets concurrently in which it presently doesn’t have any existing interest. While it’s developing unified computing systems for the enterprise, Sun’s x86 servers would put it directly into the midrange server market. Sun’s SAN (storage area network) gear would lessen its dependence on partner EMC for storage products. And Sun’s identity management and directories would complement Cisco’s existing security offerings.

Financially speaking, Cisco can better afford a big fish like Sun than IBM can. Cisco is sitting on the largest cash reserve in the technology market, now totaling close to $30 billion (£20.5 billion). While IBM is among the top 10 wealthiest tech companies, its cash reserves total less than $10 billion (£6.8 billion). For IBM to buy Sun would consume nearly three-quarters of its cash (in an all-cash deal), while Cisco could snap Sun up for less than one-quarter of its reserve.

A Sun acquisition by either company would have mixed implications for channel partners. Sun has the smallest channel, compared with Cisco’s more than 17,000 reseller partners and IBM’s 100,000 business partners. Injecting Sun’s existing 3,000 partners into either of those big ponds would increase competition for sales and tech support resources. However, those same resellers would gain access to a whole breadth of products that are currently unavailable in the exclusive Sun channel.

Neither IBM nor Sun is commenting on the potential acquisition, and Cisco certainly isn’t offering any insights. However, Cisco has published its criteria for acquisitions, and the key driver is access to new markets. While one insider said Cisco is staying out of this since it probably doesn’t want to enter a market it doesn’t lead, acquiring Sun would certainly give it access to new markets.

It may be days—if not weeks—before rumours and leaks about IBM and Sun are laid to rest. And many industry analysts and pundits will dissect and examine the anatomy of this alleged deal. Whether or not this proposed acquisition is founded in reality, the IBM-Sun news is likely signaling the beginning of major consolidation in the technology industry that will ripple through the channel and into the end-user data centre.