Here’s a thought; put a Data Center server in your house. Wild? How about putting Data Center Servers in a building to heat the building? Still wild?
Well, one of the big problems that environmentalists have been aware of are the waste of energy that comes with modern electrical systems. Granted you need the energy to run operations, but is there a way that all the energy might be used?
Consider that cloud computing is hot, literally. That means that the electricity consumed by computers and other IT equipment has been growing rapidly in recent years, and it has become a substantial part of the global energy market.
Cloud Data Centers
Data Centers have thousands of servers in their location, which means among other things that they generate a lot of heat. So could there be a way to harness the servers in a way that they still provide data center operations, but not be as toxic to the environment?
That question was recently the subject of a white paper by Microsoft engineers, The Data Furnace: Heating Up with Cloud Computing.
Consider the question. Do data center servers need to be at a data center? Is this the only location where the work can be done? In the white paper, they present the notion that servers can be sent to homes and of?ce buildings and used as a primary heat source. They call this approach the “Data Furnace or DF”.
There are three advantages that Data Furnaces have over traditional data centers. They have a smaller carbon footprint; there is a reduced total cost of ownership per server, and they can reside in closer proximity to users. From the homeowner’s perspective, a DF is equivalent to a typical heating system. Here a metal cabinet is shipped to the home and then added to the ductwork or hot water pipes. By this account, if cloud-computing applications can exploit the differences in the cost structure and resource pro?le between Data Furnaces and conventional data centers then from a technical perspective, DFs can create new opportunities for both lower cost and improved quality of service.
One Solution to Energy Consumption
Even if you wanted to make this approach feasible, what are some of the steps that you would implement? Start by replacing electric resistive heating elements with silicon heating elements. This means that you would use the same electricity source to create heat and to handle computation. This would allow the IT industry to grow its computing capacity without necessarily increasing overall electricity consumption.
How could a homeowner benefit from this approach? A homeowner could agree to site-place data servers in his or her utility room. The server performs computing tasks such as scientific processing or Web crawling at certain periods of the day or during certain seasons. For the homeowner that server farm could help subsidize the heating bill, especially at night or during the winter months.
Besides the scientific technologies that make this possible there are systems management technologies, sensor networks that improve physical security, and reduced component pricing. These would help make the home-site a measurable option.
Three Types of Furnace Configurations
To make this approach viable, where data center servers could be relocated elsewhere here are some configurations that can be used.
- Seasonal ones that use low-cost servers to perform computations mainly at night or during the winter, offering some heat subsidy to the host building
- Neighborhood ones that could help improve computing services because of their geographic proximity to the users
- Urban data furnaces that would operate year-around (this is the apartment building example); these configurations would make the most sense in colder weather, much like the seasonal ones
What makes this approach intriguing is that it would change the dynamics of the data center. One problem that they have is that if the data center goes down, everything in it goes down. This furnace approach could be an alternative since by design it would have limited servers. Still security would be paramount and it would be necessary to determine just what kind of server could operate away from the central data center site. After all, not all data is equal.