What Causes Computer Latency?

by John Papiewski

Although even garden-variety PCs execute billions of calculations per second, the various components of a computer must frequently wait for tasks to finish. Computer technicians call these delays latencies. Some latencies you can see for yourself; other kinds are imperceptible but still slow the computer's performance in subtle ways. Although new technologies tend to reduce latencies, some delays invariably remain.

Memory

A computer's microprocessor depends on its random access memory to store programs and data. RAM, however, operates at a different speed than the central processing unit, so the computer must wait a few billionths of a second for memory to send data back to the CPU. As with other kinds of latencies, brief delays add up to large reductions in speed when multiplied by millions of operations. Computer chip designers alleviate some of the latency by incorporating a high-speed type of memory directly onto the processor chip. This memory, called cache memory, saves recently used data. Before the CPU requests data from RAM, it checks the cache first. Cache technology has substantially reduced the effects of memory latency, although it has not eliminated it altogether.

Hard Drive

A hard drive stores information magnetically on platters that spin at speeds up to 15,000 revolutions per minute. A mechanism moves across the platter surface, searching for data. Although it takes only a few thousandths of a second to locate information on the drive, the computer requests information from the drive frequently, turning this brief delay into a significant latency. To remedy this delay, some computers set aside large amounts of RAM memory as buffers for the hard drive; the processor can access data in RAM hundreds of times faster than in the drive itself. Another technology that has reduced hard drive wait times is the use of solid-state drives. SSDs record information in flash memory, making data access up to 70 times faster than traditional hard drives.

Internet

Internet latencies are similar to that of a hard drive -- several milliseconds -- although this varies depending on website location and Internet data traffic congestion. These latencies result from the Internet's complex layout; a request for data from your PC travels through many networks. After it reaches its destination, the response must travel all the way back to your computer. Websites located in other continents and countries take longer to respond than local ones.

Task Load

A computer can take longer to respond to your keystrokes and mouse clicks if you have several simultaneous, demanding tasks running at the same time. The microprocessor services all the programs running on your computer, so if it is busy with others, it has less time for word processing or Web browsing. Computers with multicore processors are less prone to these latencies because the cores divide the workload, making the computer faster overall.

About the Author

Chicago native J.T. Barett has a Bachelor of Science in physics from Northeastern Illinois University and has been writing since 1991. He has contributed to "Foresight Update," a nanotechnology newsletter from the Foresight Institute. He also contributed to the book, "Nanotechnology: Molecular Speculations on Global Abundance."

Photo Credits

  • photo_camera Photodisc/Photodisc/Getty Images