In-Depth
Analysis: A Closer Look at IBM's Green Software Push
If Green Software is to have an impact, the software industry as a whole needs to get onboard
For over a year, IBM Corp. has made “greening” a centerpiece of its marketing efforts, kicking off a $1 billion Green IT investment last spring ("Project Big Green"), touting mainframe-centric Greening initiatives, and -- just recently -- trumpeting a new "Green" approach to software!
There's been little, if any, reaction to Big Blue's Green IT push.
Some observers -- especially on the mainframe side -- have groused that IBM might be exploiting one issue (rising data center power and cooling costs) to obscure another -- the prohibitive cost of mainframe software (see http://esj.com/Case_Study/article.aspx?EditorialsID=2650).
For the most part, however, industry experts seem to have bought into Big Blue's Green IT push: as an attempt to highlight the legitimate eco-credentials of its hardware, software, and services offerings -- while selling more hardware, software, and services in the process (see http://esj.com/enterprise/article.aspx?EditorialsID=2856).
There are many reasons to take IBM's claims seriously: by 2020, according to a recent McKinsey & Co. study, the data center should surpass the airline industry as the biggest overall producer of greenhouse gas pollutants. Such research helps cast IBM's Green IT efforts as both prescient and responsible.
Recently, however, some industry watchers balked when IBM trumpeted the next phase of its Project Big Green: Green Software. It isn't that eco-aware software can't help make a difference, says Wayne Kernochan, a principal with InfoStructure Associates; it's that IBM's Green Software reach seems to be exceeding its grasp. If Green Software is to have an impact, the software industry as a whole -- and not just the big hardware OEMs -- needs to get onboard. In this sense, green software is a much more complicated proposition than the green hardware offerings HP, IBM, and Sun have been touting for over a year.
Kernochan concedes that Big Blue had all of its ducks in their requisite rows. IBM even cites research which claims that it takes 27 watts of hardware to support 1 watt of software. Armonk even marshaled a list of areas in which software can help drive substantive greening (e.g., virtualization, energy monitoring or metering, and travel or paper reductions).
Kernochan isn't buying it. Not completely, anyway.
"[There's] an unspoken, lingering question: Just how effective will these software solutions actually be at reducing carbon emissions or even energy consumption?" he wonders. "To put it another way, in what specific areas of the world economy can IT -- not just software -- have a real impact on carbon emissions, and does IBM's software target these areas effectively?
For example, some of IBM's Green Ideas probably won't have a substantive impact when translated into real-world situations. Consider Big Blue's suggestion that companies ratchet up the use of collaborative software such as Lotus Notes to reduce employee travel responsibilities.
It's an idea that sounds great on paper, but -- when it's put before sales managers, marketing managers, and other stakeholders for whom face-time with clients or prospective customers is an irreplaceable opportunity -- it's going to meet with resistance. While improved collaborative tools and other technologies might help improve everyday business experiences, few companies see them as outright replacements for bread-and-butter business activities such as sales calls and marketing tours.
Companies will probably embrace collaborative technologies (including video conferencing solutions) to help reduce travel expenses, and -- yes -- every pound of CO2 reduction helps, but probably not by as much as IBM expects. More to the point, Kernochan suggests, companies seem less motivated -- in this instance, at least -- by civic than by financial drivers.
"IBM's collaborative and remote-training software aims to help businesses reduce travel costs and requirements. There is no doubt that businesses are hungry for this type of assistance, but increasing air travel costs due to rising fuel prices seem to be more of an incentive than green concerns," he explains.
Kernochan says other Green Software focus areas are even more fluid, so to speak. "[For example,] reducing the use of paper -- even assuming after all these years that yet another 'computers will create the paperless office scheme' will succeed -- has only a tangential impact on … [the need to redesign] business processes to use less energy," he notes.
Kernochan points to an earlier IBM briefing (on the subject of Green IT in general) where Big Blue claimed that IT will soon account for nearly 2 percent of the world's total energy usage -- with travel accounting for another 2 percent.
The upshot, he says, is that other industries or functions account for about 96 percent of worldwide energy usage.
"In other words, to significantly reduce carbon emissions, we will need to (a) reduce IT energy output beyond simply reducing the rate of increase; (b) reduce -- not just prevent from increasing -- business and consumer travel, especially long-distance plane travel; and (c) attack other sources of carbon emissions across multiple industries," Kernochan argues.
Big Blue is doing its part, he stresses, conceding that IBM's Green Software push clearly does aim for a reduction, rather than a mere leveling off, of energy emissions. "I am not at all critical of IBM's green efforts, nor of the prominent place the company gives to software," he indicates. "The point here is not that IBM made a mistake; it is that there is another shoe to drop in green software, one which allows businesses of all stripes to attack carbon reduction more effectively in their processes."
What's needed, Kernochan argues, isn't so much a revolution in software development but a retrofitting of existing software assets. Call it the en-Greening of enterprise software.
"I suggest that there are two key 'new things' needed: first, metering and redesign of software that assesses the likely environmental effects of every aspect of a business and its business processes. Second, add-ons embedded in existing apps that measure not just prices and costs but also carbon effects. In other words, to be a real help to businesses and to have a truly global impact on carbon emissions, software needs to do what it has always done well: analyze and improve tasks and processes -- only this time its metrics should include the environment, not merely the market."
In this sense, there's only so much that IBM -- or any of the other big vendors (e.g., HP, Sun, Dell Computer Corp.) -- can do; the software industry as a whole -- and particularly the big ERP suite vendors -- has to chip in, too.
"[T]he major enterprise-application vendors are not talking much about adding carbon-reduction considerations to ERP and business-process management. In fact, only a few narrowly vertical ISVs advertise this type of green capability -- including, ironically, landscape design," he says.
"[I]t is not that IBM is doing a bad job of creating and offering effective green software, it is that virtually all enterprise application vendors are ignoring the elephant in the room: the 96 percent of energy emissions that today's green software [offerings] do not affect."