We often leave our computer switched on overnight. Why?

• Our computer provides continuous gentle heating of ≈200W, rather than the short bursts of 2kW–3kW high-powered heating typical of a pulse-modulated timer-operated central heating system; reducing the amount of unnatural variation in the temperature of our home during the night and reducing the supply/demand balancing problems of energy companies.  When the ambient temperature is just a few degrees below that required for our home to be comfortable, this effect is more comfortable than central heating.  The alternative?
  
• Rather than simply dissipating electrical energy as heat, we run useful computer programs with the electrical energy, for example,
  • virus scanners,
  • disc defragmentation & backup archival programs and
  • distributed supercomputing software such as folding@home; a medical research programme at Stanford University.

Managing the heat by-product on a larger scale

In office spaces with large numbers of personal computers and comparatively small exterior surface areas, the heating effect of computer equipment (and even from human bodies, when the space is occupied, which supply about 70W of heating each) can be particularly significant.  Every workstation behaves as though it incorporates a low-powered electrical heater (≈40W–400W).  This effect should not be ignored by building designers, and if properly managed, can be useful since it reduces winter heating requirements.  On the other hand in summer, the presence of heat-generating equipment can complicate building design by increasing requirements for cooling and ventilation.  When the distributions of heating and cooling resources are most dissimilar, occupant comfort and energy efficiency are sub-optimal.

In an ideal world, every computer would incorporate an ambient air temperature sensor so as to provide adaptive heating through temperature-based control of distributed computer software & background process workloads.  Building designers might also be given limited centralised control of such heating via the computer network, so that proportional/integral/differential control may be applied locally to each building space in a way that takes proper account of the building's heat dynamics and the total effects of all environmental control equipment at different times of day; optimising comfort for building occupants whilst minimising overall energy consumption.

Purpose-built electrical heaters with distributed computing function

In the coolest periods of winter, concentrated heat-sources are often required.  In an ideal world, all electrical energy would be usefully applied and none dissipated directly into heat.  Electrical heaters would incorporate thousands of simple microprocessors connected to the internet, helping to advance scientific research and collecting micro-payments from its commissioners, whilst at the same time heating homes and offices as desired by their occupants.

The economics of manufacturing and distributing such complex electrical heaters (either incorporating heat-tolerant microprocessors, or having unusually large heating surfaces, or some “happy medium” between the two) might be intractable right now, but there might be a business-case for doing this in the home beyond about 2025–2035.  This idea is most likely to become viable if ways can be found to salvage microprocessors for this purpose when they are no longer useful for other computing devices.

Conclusions

In an ideal world, process technology would always be situated where its by-products were most useful, and fine-tuned to produce the specific by-products most locally beneficial.  Thoughtful design, environmentally responsible legislation and conscientious consumers could take us a long way toward such an ideal world in a short period of time.

You can already contribute meaningfully toward the progress of medical research whilst gaining a net personal benefit (more comfortable heating in spring and autumn), by installing a small computer program from the folding@home web page.