Understanding Two Approaches to a Computer’s Useful Life: Planned Obsolescence and Time Until Failure

Article by John Garger (20,844 pts ) , published Jul 29, 2009

There are two answers to the question: How long should my computer last? One deals with the obvious answer of computer failure. The other is rooted in planned obsolescence.

Computers are made up of many parts made from different manufacturers. Standards, some imposed on the manufacturers by outside entities and some self-imposed, ensure that the parts that are supposed to work together do so. However, without the assistance of a systems integrator, some computers parts just naturally last longer than others do.

Under normal use, a computer’s case could conceivably last many decades. The power supply, however, inherently is a component doomed to failure in the long run. No electrical component can last forever due mainly to the stresses of constant heating and cooling caused by friction from the electricity traveling through its wires, capacitors, and other parts. In fact, next to memory failure, power supplies are the most likely component to go on a computer.

To answer the question of how long a computer should last we must first adopt a paradigm. Are we talking about computer failure or how long the computer remains useful? One could argue that a computer from the late 1970s may not have yet physically failed but that it has failed to be useful juxtaposed to a modern PC. In a sense, the old computer has not “lasted” because it no longer provides its user with an advantage over abandoning it for a more modern computer. Both types of computer longevity are briefly discussed below.

Physical Failure as a Measure of Computer Longevity

Certainly, when a computer ceases to function it is no longer useful to its owner. From the time the computer was first turned on to the day it failed to function properly is a measure of how long the computer lasted. As mentioned above, the modern computer is made up of many parts. The failure of a primary component such as the CPU, memory, motherboard, etc. means the usefulness of the computer has ceased. Failure of a minor component such as DVD drive, USB port, or internal fan simply means that the computer’s usefulness has decreased but not necessarily to zero. Companies, when deploying hundreds and thousands of computers in an organization, are constantly calculating the productivity losses due to these two types of failures.

Some manufacturers place failure estimates on their products that estimate how long the component will last until its first failure. Often printed on hard drives, the Mean Time Between Failures (MTBF) is just such as estimate. A MTBF of 30,000 hours means that under normal load, the hard drive can be expected to fail at its 30,000th hour of operation. Of course, this is just an estimate based on statistical operations and it is only an average. While some hard drives with a MTBF of 30,000 hours will fail before this time, some will continue to operate far beyond.

Computers especially are sensitive to changes in ambient heat, humidity, and power spikes. Some components in the system attempt to control these factors but users typically put their computers through far more situations than a component manufacturer can simulate in a lab. The result is no way of knowing exactly how long a computer will last.

The best way to guard against unexpected computer failure is to do two things. First, do regular backups of pertinent data. Automated backups are best since they require no human intervention. How often to back up is a question of data sensitivity. If you absolutely cannot lose your data from a previous day, daily backups are recommended. Second, make sure you have a backup computer where you can access your backed up files. It may take days or weeks to fix a computer, especially if the computer is under warranty by a large corporation. The need to procure Return Merchandise Authorization (RMA) numbers, diagnose, and repair the computer can be time consuming and frustrating when needed files cannot be accessed because of routine hardware failure.