Dell’s recall of 4 million Sony-built batteries – the largest electronics product recall in history – and the subsequent Internet footage of laptops bursting into flames, just confirmed what every travelling executive knows: not only are the lithium-ion batteries used in laptops underpowered, but the heat they generate makes them (literally) too hot to handle.
The problem is that batteries are behind the technology curve. Although other laptop components, such as processors and memory, have advanced at astonishing speed in recent years, improvements in battery performance have been, comparatively, much more modest. Moreover, as users have become hungry for smaller, neater – but still-powerful – form factors, mobile device manufacturers have struggled to find batteries that can combine miniaturisation and longevity.
The problem goes back to use of lithium-ion technologies in rechargeable batteries. That may have been a revolution, says Jim Tully, chief researcher at IT advisory group Gartner, but further advances has been slow.
The technology, which generates current through the migration of ions between the battery’s electrodes, offers four times the energy density of traditional lead-acid batteries, at relatively low cost. There have been some incremental improvements in capacity through the use of nanotechnologies; indeed, researchers at Altair Nanotechnologies, for example, have manipulated electrodes at microscopic levels to massively increase their surface area, thereby improving performance and lifespan.
But, says Tully, the real gains are now being made by developers of other energy sources. Electronics giant Toshiba has produced a super-charging battery that can regain 80% of its capacity in less than a minute. However, developing a stable commercial version of that battery has proved challenging, and previous promises of a 2006 release are “still under consideration”.
Meanwhile, another alternatives technology – micro-fuel cells are starting to look like a plausible option. Methanol-based micro-cells have been shown to provide up to 10 times the storage capacity of a lithium battery, and can easily be recharged simply by adding a new methanol cartridge.
Livermore, California-based UltraCell, for example, has already held demonstrations of its hot-swappable fuel cell units, which are capable of running a laptop for several days. Its technology is currently being trialled by the US Army, where long-life and portability are highly desirable.
However, fuel cells are ill-equipped to provide power to many portable devices, says Gartner’s Tully. Devices, such as mobile phones, draw power at uneven rates; fuel cells, however, provide a constant supply, but are poor at dealing with such surges.
As a result, it is likely that fuel cells, when eventually employed, will be used in conjunction with some other energy store, effectively supplying back-up power, he adds.
This concept of a back-up store to a main power source is likely to become increasingly popular, and could include other methods of power generation, including solar panels, closed-loop induction and facilities to “scavenge energy” from the body heat of the device. Some of these technologies are still several years from any kind of large-scale roll-out, but they promise to herald a new, longer-lasting mobile experience. Whether they will also be a source of pyrotechnics has yet to be seen.
Direct power
Direct current could slash data centre operating costs.
As little as 30% of the power that enters a data centre is actually used to drive microprocessors. A lot of this ‘excess’ power is used to drive the fans that keep the chips cools, but a significant portion is simply lost to inefficient electrical systems, and in particular to wasteful power supplies.
This issue is already partially addressed by blade and rack system vendors who are building enclosures that enable servers to share communal power supplies. However, some larger data centre operators are now also experimenting with delivering electricity directly to servers as direct current (DC).
In conventional data centres, electricity arrives at the users’ premises as alternating current (AC) and then undergoes a process of transformations to direct current. This is because computers require unidirectional DC to function, but AC is safer to work with, and makes it easier and cheaper to transport electricity from point to point. However, each conversion process dissipates power that could otherwise be used to drive machines; this extra heat is most unwelcome.
DC proponents say this wasted energy can be reclaimed by dispensing with AC distribution systems – reducing wastage, and in the process dispensing entirely with the need for computer power supplies. This can reduce power consumption by 20% in working environments, and by 50% in systems that experience a lot of idle time – a potentially significant saving for those with big power bills.
But, DC has drawbacks. DC cabling is several times thicker and more expensive than familiar AC cables, and significantly more difficult to pull through narrow ducts and to bend around awkward corners. Once installed, it also takes special skills to maintain it safely. In all but the biggest data centres these drawbacks will probably defeat the case for DC, but for green-field sites, DC distribution may prove to be the way forward.