POWER > Business Strategy > ROI

IBM’s Steve Mills on the Economics of IT and Cost-Saving Strategies

Steve Mills senior VP and group executive, IBM software and Systems — Photo by Bill Bernstein

If you need an expert on the economics of IT, there’s no one better than Steve Mills, senior vice president and group executive, IBM Software and Systems. Mills has been with IBM since 1973, steadily gaining more responsibilities and bringing him to his current position where he sets the direction for more than 100,000 employees spanning software and hardware development, manufacturing, sales, marketing and support.

In an industry that offers so many choices and often many mixed messages, it can be difficult to know what IT decisions will deliver the best economic advantage. Mills is focused on helping IBM clients get the most value from their IT spending. IBM Systems Magazine sat down with him to learn more about the economics of IT.

Q: What value does IT bring to business?
IT is all about the automation and optimization around business processes—the things companies must do to run their businesses as well as using IT as a strategic tool to provide better customer service, faster customer service and develop more insight about the markets the companies compete in. It represents a strategic advantage when used effectively to power a business both on an everyday basis as well as on a long-term forward-looking basis.

Q: What IT operating costs are rising? Are any decreasing?
Hardware costs are going down. The hardware cost per unit of processing capacity has been dropping basically every year. The phenomenon of Moore’s law, where price/performance doubles roughly every 12 to 18 months, has continued in all kinds of forms whether it’s server capacity, server performance, storage capacity or the cost of storage. Networking doesn’t generally come down quite as fast, but it’s coming down at a pretty good clip, so the hardware side of the business has been dropping in price. Of course, what that implies is for every dollar you spend on software, you’re going to be able to process more. So, generally, software prices or software budgets are not falling but what you get for the money certainly is rising.

How You Use It Intel servers often have only 10-20% utilization UNIX systems operate at about 2x that Power Systems servers can sustain 70% utilization

Labor costs continue to be the biggest single issue. Labor also accounts for the fastest growing element of IT. Businesses around the world spend considerably more on labor than they do on the combination of hardware and software, so labor costs are much, much higher than the asset costs and continue to expand and grow. Beyond that, you have physical facilities, so as businesses end up with a lot of IT, the cost of floor space, power and cooling are also rising. When businesses look at their IT budgets today versus a decade ago, it’s quite startling in terms of how the monies have shifted away from the individual cost of the hardware box and much more toward all of the surrounding costs.

Q: What strategies can businesses use to manage cost?
I think you have to have a comprehensive approach to optimizing the way in which you use IT. Underutilized IT assets are a problem. Server utilization in the industry today still is well below what it could ideally be. There are 33 million servers deployed in the world. When talking about this sprawl, you must consider the proliferation of Intel* technology-based infrastructure. They are the most poorly loaded systems deployed today. Often they have only 10 to 20 percent utilization. UNIX* systems tend to be operating at twice that. IBM Power Systems* servers are certainly capable of sustaining 70 percent utilization, which in the UNIX world at least, is unheard of from a historic perspective. Mainframes are often run at 80-plus percent. We see customers that peak their mainframes at nearly 100 percent CPU utilization on and off throughout the processing day—really efficient use of the systems through effective loading of the boxes. That means well-structured applications that can be virtualized and shared, the effective use of logical partitioning, system balancing in terms of memory and I/O. All of those things come into play. If you design and do it well, you could end up with a smaller, more compact, more compressed footprint—and that saves you money.

Now beyond that, it’s a focus on not having redundant application assets. Businesses around the world often have more than one of everything. We see financial services companies that might have eight, 10 or 12 collection systems. Insurance companies may have multiple ways to do the origination of policies. They may have multiple claims processing systems. Every industry has its examples of duplicates. Those duplicates are the biggest single cost driver. Not only do they chew up processing capacity, capital costs and IT operational labor but they also actually impact the overall operational cost of the business. The biggest savings are always derived from reducing redundant application assets and getting to ideally one and only one instance of each and every application function based upon the processes of a company. That is where the big savings lie.

Tami Deedrick is the former managing editor of IBM Systems Magazine, Power Systems edition.

comments powered by Disqus



2019 Solutions Edition

A Comprehensive Online Buyer's Guide to Solutions, Services and Education.

IBM Systems Magazine Subscribe Box Read Now Link Subscribe Now Link iPad App Google Play Store