As many of you will have read, IBM has agreed, in principle, to sell its x86 server business (System-x, Blade Center and Flex-System) to Chinese PC maker Lenovo. (Note: Flex-System is the twin brother to Hitachi’s Compute Blade 500 product line). This is a deal that has been speculated about for months, but what does this say about the state of the server market, and what does it mean for new server and solution customers?
The Big Squeeze
It has been many years since IBM re-invented itself as a business focused on the high margin businesses of services, consulting and software – in comparison with which, hardware – and particularly the lower end of the hardware business – has looked less and less appealing. In 2005 they sold the PC and laptop business to Lenovo, and so, with that relationship already in place it seemed natural that in time the related low end server business would eventually follow, as the growth of large scale-out data centers, and the erosion of differentiation, continued to compress margins in this market.
As a result of these trends, we are seeing an increasing separation between companies who focus on business value around solutions and applications, and those whose expertise and success is all about optimizing manufacturing and distribution costs based on volume.
By making this move now, IBM has indicated that it believes that this part of its business reached the tipping point where low cost is a bigger market driver than added value.
One notable feature of this announcement is that IBM retained its mainframe and high-end UNIX hardware businesses. It’s long been understood that these markets are in a slow inexorable decline, but nevertheless maintain high relevance to a significant segment of the market who appreciate and value the high-end features that these systems provide. Of course, the key here is differentiation. By meeting the demanding needs of customers with mission-critical applications and delivering differentiated value, the company can keep margins high and match the expectations from its other businesses.
Lenovo: A New Enterprise Competitor?
Clearly - and by design - Lenovo is a horse of a different color. This is a company which has demonstrated success in the PC and smartphone market by cutting costs though cheaper manufacturing and efficient distribution. However, even at the low end, the enterprise server business is not just about creating the cheapest product. Lenovo will clearly have to move outside its traditional comfort zone to succeed. It has no experience building high end products to meet the needs of exacting enterprise customers and has yet to establish strong relationships with critical software makers such as VMware, Oracle or SAP that drive customer purchases in this space.
It also remains to be seen how large enterprise customers and government agencies, particularly in the US and Europe, will feel about transitioning to a supplier based in China.
HDS: The best of Both Worlds
What if you could "have your cake and eat it too", (as the old saying goes)? Imagine you could have a product with all the price advantages of commodity servers based on Intel processors, but with the kind of high-end features that you would expect from proprietary UNIX servers, or even mainframes: features such as built-in logical partitioning, scalable symmetric multi-processing and automated failure identification and recovery. Finally, imagine that this product was tightly integrated via management software into high value solutions with the industry's leading enterprise applications such as SAP and VMware. Such a product would offer tremendous value, right?
Of course, this is not simply an exercise in imagination - what I'm describing is the real Hitachi Compute Blade family of servers which power our Unified Compute Platform (UCP) solutions.
In addition, HDS is an American company, with a Japanese parent, which may be important for some.
So what should System X, Bladecenter and Flex System customers do now?
IBM's existing customers are understandably nervous. No doubt they are currently asking about plans for ongoing support and demanding technology roadmaps for the parts of these product lines that are retained by IBM. Will there be a divergence? Could there be compatibility issues?
In many ways IBM's announcement is an implicit endorsement of HDS’s strategy to focus on solutions and to deliver high value for enterprise customers. At HDS we see servers as a critical component of our solutions - one which allows us to build additional value into the whole product. However, because our servers have high value differentiated features and we don't try to sell them as a commodity product at high volume, they remain a sustainable business - and something we are committed to for the long term.
Now is a great time for IBM's x86 platform customers to take a look at HDS' UCP offerings based on the Hitachi Compute Blade family, and compare. I think they will like what they see, and find that there's a viable alternative on the market today that is not only mature and reliable but comes from a company which is 100% focused on the needs of customers like themselves.