Access time

from Wikipedia, the free encyclopedia

In addition to the data transfer rate, the access time is an important performance measure of storage drives such as hard drives or optical drives and other storage media such as main memory . It describes how quickly the relevant storage medium can carry out the relevant process or provide the relevant data after receiving a write or read command.

Volatile memory

Processor register

The processor registers of a computer have the fastest access time , as these are usually operated with the clock frequency of the processor . There is only a very limited number of these registers; their access time is well below a nanosecond in modern processors .

Processor cache

A computer's processor cache has the next fastest access time . This is usually implemented as SRAM and can be located both inside the processor and on the motherboard (but then in close proximity to the processor). Its access time is in the range of a few nanoseconds.

random access memory

The main memory of a computer is generally considerably larger than the processor cache and is therefore made from DRAM that is inexpensive to produce . The access time to the main memory is higher (around 60–70 nanoseconds) than the access time to the processor cache. So-called RAM disks can be set up in the main memory, which are addressed like a hard disk, but have the access time (and the transience) of the main memory.

Non-volatile memory

Flash memory

In the case of flash memory , such as B. is used in memory cards , USB sticks and increasingly also in solid state drives, there is no need for a mechanical movement of a read / write head. Such memories therefore have considerably shorter access times (around 250 µs) than hard disks or optical drives, but at higher acquisition costs.

Hard drives and optical drives

In the case of rotating media such as hard drives or optical media such as CDs / DVDs, the access time is made up of the time required to position the read / write head over the track (also known as seek time ) and the latency that elapses until the correct sector appears on the medium under the head ( latency ). In the case of a mostly assumed random distribution of the accesses, it only makes sense to specify an average access time in which the parameters (previous head position and sector position) are evenly distributed.

Modern hard disk drives achieve average access times of around 9 ms , depending on the number of revolutions (typically for 7,200 revolutions). SATA drives have mechanisms that allow the access commands arriving on the drive to be rearranged to shorten the access time ( native command queuing ).

The access time of optical drives is slower than that of hard drives.

Floppy drives are, among other things, significantly slower due to their slow and comparatively cumbersome mechanics.

Tape drives

Tape drives generally have a very long access time (several seconds or longer) due to the need to fast forward or rewind the tape. Tape drives are therefore mainly used for archiving purposes.

literature

  • Dietrich Franz: Basics of EDP. Gabler Verlag, Wiesbaden 1991, ISBN 3-409-00164-6 .
  • Kurt Bauknecht, Carl August Zehnder: Basics for the use of IT. 5th, completely revised and expanded edition. BG Teubner Verlag, Stuttgart 1997, ISBN 3-519-42450-9 .
  • Kurt Bauknecht, Carl August Zehnder: Basics of data processing. Methods and concepts for the applications. 4th edition. BG Teubner Verlag, Stuttgart 1989, ISBN 3-519-32450-4 .
  • Klaus Fischer: Basics of automated data processing. 2nd Edition. Gabler Verlag, Wiesbaden 1986, ISBN 3-409-02155-8 .
  • H. Schecher: Functional structure of digital computing systems. Springer Verlag, Berlin / Heidelberg 1973, ISBN 3-540-06275-0 .
  • Carl Schneider: Data processing lexicon. 2nd Edition. Gabler Verlag, Wiesbaden 1976, ISBN 3-409-31834-8 .
  • Manfred Broy, Otto Spaniol (Hrsg.): VDI-Lexikon Informatik und Kommunikationstechnik. 2nd Edition. Springer Verlag, Berlin / Heidelberg 1999, ISBN 3-642-46846-2 .

Web links