
As we expected, 2008 was an important year for power management, with desktop power management suppliers winning large-scale deals and enjoying high growth rates. In the datacenter, however, power management made little commercial progress: complexity, concerns over reliability, and a weak return on investment in some cases, prevented widespread adoption.
In 2009, we expect enterprise desktop power management software to continue enjoying rapid adoption – although ROI is usually less than six months, market penetration is surprisingly low in all geographies. We expect 1E, the UK software company to consolidate its position as the leading supplier in 2009, following some big deals with AT&T and Microsoft in 2008. Rivals Verdiem, Faronics and BigFix will also continue to grow, but their market shares and ability to win big deals will depend on their execution and marketing. However, as the technology is adopted more widely, and as the leading suppliers fight for market share, prices per desktop are likely to fall during 2009.
The technology of desktop power management is now fairly stable, and progress in this area will mostly relate to ease of installation and use, improved analytics, increased flexibility of policy controls, and new packaging for different types and sizes of businesses. 1E, for example, is seeking to expand into the SME market, where penetration of centralized power management tool is almost nonexistent. We do not believe efforts to promote consumer power management products, even as freeware, are likely to prove important – operating system power management is already sufficient. In the datacenter, we eventually expect power management to become much more widely used than it is now. (We have seen surveys that put datacenter power management use at 10% or higher, but this seems unlikely. In most datacenters, automated server power management is not used at all.)
However, we do not expect 2009 to be a breakthrough year: Server power management is still technically immature, and this is not an area for risk taking. In 2009, there will be many server and datacenter power management pilots and trails undertaken, but few customers will commit to the technology on an enterprise scale. Suppliers such as Cassatt, Power Assure and Alpiron will still find it hard to convince CIOS that the technology is ready for the mainstream. Part of the reason for this is that most datacenter managers have yet to properly ‘wire up’ their datacenters, putting in the sensing and reporting systems that underpin the next stage – control and automation (see datacenter section, below).
We do, however, expect to see a lot of innovation in datacenter and server power management in 2009, with 1E emerging as an innovative new entrant. It has developed new ‘sleepy server’ technology that involves turning machines down, rather than off, and this may help turn the market. 1E is beginning to build strong relationships and customers in very big companies, some of which are very likely to trial the technology. In power management, Hewlett-Packard and IBM have so far focused more on power capping than on power management, and this technology has become increasingly sophisticated.
We expect that both IBM and HP will introduce more sophisticated server of datacenter-wide power management software, possibly in conjunction with a virtualization software supplier (VMware being the most likely) although the timescales and positioning are not clear. CA Inc is also likely to more formally productize some of its capabilities in this area.
Main changes to the market expected this year:
Companies at the forefront of disruption: