IBM-Lenovo Server Agreement Basically A Done Deal

A few weeks ago, U.S. officials OK’d the way for Lenovo to acquire IBM’s x86-based server business for $2.3B. While other countries still need to approve the deal, the U.S. Treasury Department’s CFIUS (Committee on Foreign Investment in the United States) approval was the last real hurdle Lenovo had to overcome.  This is not to say another country’s approval isn’t important or challenging, but it’s expected to pass other places, as the U.S. is known to have the most stringent policies, particularly with the recent icy relations between the U.S. and China.

Lenovo is mum on exactly which approvals need to come next and issued the statement, “Out of respect for the confidentiality of the process, we will not have any further comment until we have met all conditions for closure.” Lenovo is saying they expect the deal to close by the end of the year, four months from now.

Timothy Prickett-Morgan at EnterpriseTech had a little more to say about it, saying, “In addition to the CFIUS approval in the United States, the IBM-Lenovo deal needs the approval of regulators in China, the European Commission, and Canada. Europe and China had previously approved the deal.”


So my head is already down the road assuming the deal will be approved. I believe Lenovo and IBM have already made concessions and I’m expecting them to make more down the line.

What I find so interesting is just how many people are banking on non-approval or just assuming that the go-forward Lenovo/IBM server business will be a flop. To get the full picture, I really think people need to consider the following:

  • Lenovo is buying a complete business: Unlike other acquisitions that take “piece-parts”, Lenovo is acquiring the entire IBM x86 server business. Functional groups include product development, sales, service and support, 34 R&D Labs and offices and 7 manufacturing plants. Also, in addition to vanilla racks, Lenovo also picks up blade servers, modular system for HPC (high performance computing), and networking. Complete businesses are a lot easier to integrate than pieces.
  • IBM will support servers for five years: While Lenovo is buying an in-tact business, IBM will still be supporting IBM-branded servers for a minimum of 5 years. Some are forgetting this fact and it’s an important one to enterprises.
  • Servers are margin accretive to PCs: While IBM may characterize x86 servers as a lousy business, it could be a really good business for Lenovo. Remember, Lenovo excels in PCs, tablets and phones where margins are brutal, a business Lenovo says they “love”. Anything above the PC business is just gravy.
  • IBM is still the #3 server vendor: While the IBM x86 server line has lost a bit of the luster it had a decade ago when it was driving higher unit market share, it still is a major unit volume player in servers. This matters as objects in motion are harder to divert off their current course and of course improves buying power.
  • Minimal server market overlap: Lenovo already has a server business. It’s much smaller than IBMs but it’s growing a lot faster. Lenovo is currently a major player in China- #1 market share in dense servers for large cloud players and #1 market share in towers for small businesses and branch offices. Since 2012, they’ve grown 400% in the U.S., double the nearest competitor. IBM, on the other hand, plays in classic large enterprises of business, governments, and universities. Both server businesses have low degrees of market overlap. That means less complexity and easier integration.

I think the market is unwise to jump to any immediate, negative conclusions about Lenovo’s future in the datacenter or even small business servers without considering the points above. I remember the doubters when Lenovo bought the IBM PC business as I was one of them. I remember the reaction about their foray into tablets and phones when Apple and Samsung were laying waste to the tablet and smartphone market, too. While history doesn’t always repeat itself, it’s not smart to bet everything against it, either.