Page 1 of 8
By Michael Neubarth
Once again, the IBM mainframe is at the center of a debate over its survival. The concern stems from a steep and steady decline of mainframe revenues over the past five quarters.
As network world.com reported on January 21, 2010, "IBM mainframe woes continue with big Q4 drop." For its fourth-quarter 2009 and full-year results, announced on January 19, 2010, IBM said its revenues from System z mainframe server products decreased 27 percent compared with the year-ago period, and that total delivery of System z computing power, which is measured in MIPS (millions of instructions per second), decreased 19 percent.
The dip in fourth-quarter 2009 results followed declines of 26 percent in the third quarter, 39 percent in the second quarter, and 19 percent in the first quarter.
IBM lumps together all its server revenues, so analysts must estimate how much IBM derives from its mainframe line. In analyzing the 2009 results, Timothy Prickett Morgan on itjungle.com on February 1, 2010, said, "I reckon that IBM's mainframe revenues declined by 41.5 percent to $2.55 billion."
Based on Prickett Morgan's calculations, a quarter-by-quarter comparison of System z revenues in 2008 vs. 2009 shows:
The revenue declines have sparked a new round of speculation about whether the mainframe is headed for extinction. For example, wrote Steve Hamm in Business Week on August 6, 2009: "IBM's 45-year-old line of mainframe computers has survived the onslaught of minicomputers in the 1980s, the Unix operating system in the '90s, PCs and the Internet. But a 39 percent plunge in mainframe revenues in IBM's second quarter seems to signal that this longtime mainstay of IBM's business is on its way to the junkyard."
However, saying the mainframe is dying because of the declining revenue numbers is not a straightforward deduction that can be made. As Jeffrey Clark notes in "What's Behind Declining IBM Mainframe Revenue?" on datacenterjornal.com on January 27, 2010: "A number of factors have an influence on mainframe sales; in light of these factors, these gloomy numbers may or may not indicate an overall decline in the market for mainframes in general and IBM mainframes in particular."
As Clark and other analysts observe, among the key factors to consider are the depressed state of the economy and the general pattern of IBM's product release cycles. As analysts point out, an eventual decline is natural after a new version is introduced, and before the next version arrives.
IBM introduced its last new model, the z10, two years ago, which caused a spike in mainframe revenues in 2008. The mainframe was hailed as a rejuvenated success. Then came five successive quarters of decline. The recessionary economy, coupled with the fact that IBM is set to introduce a new model in 2010, may have exacerbated a cycle in which customers tend to withhold spending while waiting for the new technology.
And indeed, this is the account that IBM endorses to explain the revenue decline. In announcing the yearly results, IBM CFO Mark Loughridge was widely quoted as saying that IBM was going to release the next generation System z later in 2010, and that the revenue decline "is consistent with what you would expect at this point in the product cycle."
However, as Jon Brodkin on networkworld.com on January 21, 2010, noted: "IBM unveiled the System z10 mainframe almost two years ago, leading to a 12.5% increase in revenue in 2008. The initial boost in sales caused by a mainframe upgrade is typically followed by a decline, but the revenue drops in 2009 were quite a bit larger than the amount gained in 2008."