IBM announced its corporate financial results for the second quarter of 2014. IBM’s mainframe business continues to impress amidst an otherwise gloomy global server business.
I should caution, as I always do, that IBM only breaks out the hardware-related revenues for its System z mainframe business in its quarterly earning reports. Thus we only get an imperfect snapshot of the overall mainframe business. IBM’s mainframe-related business also consists of software, services, and maintenance. While these other sources of IBM revenue and profit are not reported separately, it’s a safe bet they’re much larger than the direct hardware-related figures. In turn, IBM represents only a fraction of the total direct and indirect mainframe business. Many other software vendors, service providers, and other IT firms comprise the mainframe ecosystem. It’s also important not to focus too heavily on a single quarter or even a single calendar year. The mainframe business has been running for at least half a century.
With those caveats restated, IBM reported that its System z hardware revenues were down a mere 1% year over year, and capacity deliveries (“MIPS”) were flat. Is flat financial performance something to celebrate? You bet. As IBM noted, these results are 7 quarters deep into the current mainframe product cycle. In short, these are damn impressive numbers. Once again the mainframe outperformed IBM’s other hardware segments. It’s highly likely these results mean that the IBM mainframe gained server revenue marketshare. Again.
IBM also tends to enjoy better profits after the first few quarters of each mainframe product cycle. Initial sales in a product cycle are more physical, with IBM shipping more equipment than in later quarters. Much like software, when you ship fewer physical components but still collect revenues, profit margins tend to grow. As I’ve pointed out before, profits fuel innovation. I continue to predict that the mainframe will continue to outpace the rest of the server market in terms of new capabilities and new innovations. I’d even predict the gap will widen over time. For example, IBM just announced another $3 billion worth of investments in semiconductor-related research and development, and much of that R&D will find its way into future mainframes.
So why are mainframes so popular? I think IBM’s CFO makes the case quite well. The overall trends in the IT industry are swinging back toward firm ground for IBM’s mainframes. Businesses and governments are confronting ever-increasing information processing demands as the volume of data continues to surge with no end in sight. Their customers and other users are demanding more and more responsiveness, or they’ll quickly take their business elsewhere. Logical application and information system architectures continue to increase in complexity, and that complexity tends to foster infrastructure simplicity (i.e. mainframes). Trust, security, and integrity are more critical than ever to more businesses and governments. Connected computing continues to increase its share over disconnected computing, and thus centralized computing models continue to gain ground. Efficiency and multi-tenancy continue to grow in importance throughout the industry. All these factors contribute to a resilient, robust mainframe business on a long-run basis. That’s not to say that IBM can ever rest easy, but I don’t think IBM is.