====================================================================== Title: Notes on Metric System Measurement Terms Used in PCs ---------------------------------------------------------------------- Filename: METRIC.TXT WWW URL: http://www.firmware.com/support/bios/metric.htm FTP URL: ftp://ftp.firmware.com/text/metric.txt Revision: 07/06/98 TLS Micro Firmware Technical Support Summary: Defines and explains computer-related measurement terms and unit symbols, such as megabyte, kbps, MHz. ====================================================================== . Metric System - Prefixes, Unit Symbols, Powers of Ten, Multipliers deka (da) = 10 ^ 1 = 10 ten hekto (h) = 10 ^ 2 = 100 hundred kilo (k) = 10 ^ 3 = 1,000 thousand mega (M) = 10 ^ 6 = 1,000,000 million giga (G) = 10 ^ 9 = 1,000,000,000 billion tera (T) = 10 ^ 12 = 1,000,000,000,000 trillion peta (P) = 10 ^ 15 = 1,000,000,000,000,000 quadrillion exa (E) = 10 ^ 18 = 1,000,000,000,000,000,000 quintillion zetta (Z) = 10 ^ 21 = 1,000,000,000,000,000,000,000 sextillion yotta (Y) = 10 ^ 24 = 1,000,000,000,000,000,000,000,000 septillion deci (d) = 10 ^ -1 = .1 tenth centi (c) = 10 ^ -2 = .01 hundredth milli (m) = 10 ^ -3 = .001 thousandth micro (u*)= 10 ^ -6 = .000001 millionth nano (n) = 10 ^ -9 = .000000001 billionth pico (p) = 10 ^ -12 = .000000000001 trillionth femto (f) = 10 ^ -15 = .000000000000001 quadrillionth atto (a) = 10 ^ -18 = .000000000000000001 quintillionth zepto (z) = 10 ^ -21 = .000000000000000000001 sextillionth yocto (y) = 10 ^ -24 = .000000000000000000000001 septillionth * The actual unit symbol for micro is the lowercase greek symbol for mu, which looks like a u. All above prefixes and all words formed with these prefixes properly begin with lower case letters. For kilo and smaller, all unit symbols properly begin with lowercase letters. For mega and higher, all unit symbols begin with uppercase. Quantity values are correctly written with a space between the unit symbol and the quantity - 10 KB, not 10KB. There are some exceptions to definitions of these prefixes and to the above capitalization rules in the computer industry. This is explained below in the section on memory size measurement. All of the above prefixes are actually from the SI system, which is basically the replacement for the metric system. Some of the above prefixes such as yocto may actually not have been defined before SI and so may not technically be called metric prefixes. Since most people are more familiar with the term metric than with SI, and since its basically all the same, I have used the term metric throughout this document. It should also be pointed out that terms such as million and billion may be defined differently in different countries. ====================================================================== Metric System Measurement Terms Used in the Computer Industry Some uses for metric system prefixes in the computer industry include terms for measuring memory sizes, data storage capacity, data transmission bandwidth, processor speed, and ratings of electronic components. ---------------------------------------------------------------------- Memory Size and Data Storage Measurement Since computers work with the binary number system, some values in powers of 2 are represented by the closest equivalent values in the decimal metric system. This is mainly of concern with measuring memory sizes and data storage capacity. Memory sizes in computers are always expressed in binary terms (KB = 1024 bytes). Note that these are not standard SI measurements. SI prefixes are strictly decimal multipliers. The binary forms of kilobytes and megabytes have become standard throughout the computer industry, although they are incorrect uses of the SI prefixes. The IEC and ISO, together with commitees from the IEEE have done some work on proposals for some new prefixes that could be used for binary forms of measurement in the computer world. It may be awhile before these are made official and even longer before they begin to be widely used. Some information on this can be found at this URL: http://physics.nist.gov/cuu/Units/binary.html Data storage capacity (such as on hard drives) is expressed both in binary and decimal megabytes, which can be very confusing. Drive manufacturers normally advertise drive capacity in decimal megabytes. Newer BIOSes will usually show drive capacity in CMOS setup using decimal megabytes, older BIOSes use binary megabytes. DOS FDISK shows drive size in binary megabytes. Kilobyte (KB or K) = 1024 bytes (2 ^ 10) "binary kilobyte" kilobyte (kB) = 1000 bytes (10 ^ 3) "decimal kilobyte" Kilobit (Kb) = 1024 bits (2 ^ 10) "binary kilobit" kilobit (kb) = 1000 bits (10 ^ 3) "decimal kilobit" megabyte (MB or M) = 1,048,576 bytes (2 ^ 20) megabyte (MB) = 1,000,000 bytes (10 ^ 6) megabit (Mb) = 1,048,576 bits (2 ^ 20) megabit (Mb) = 1,000,000 bits (10 ^ 6) gigabyte (GB or G) = 1,073,741,824 bytes (2 ^ 30) gigabyte (GB) = 1,000,000,000 bytes (10 ^ 9) gigabit (Gb) = 1,073,741,824 bits (2 ^ 30) gigabit (Gb) = 1,000,000,000 bits (10 ^ 9) terabyte (TB) = 1,099,511,627,776 bytes (2 ^ 40) petabyte (PB) = 1,125,899,906,842,620 bytes (2 ^ 50) There is widespread confusion regarding the difference between binary and decimal kilobytes and megabytes. Since both types of measurements are used when referring to data storage capacity, it is important to distinguish between the two. Since the abbreviation for kilo is a lowercase k, this is used for decimal kilobytes and an uppercase K is used to indicate binary kilobytes. This is an established convention and should always be followed (though it isn't always). Beyond that, things get a little fuzzy. Since an upper case M is always used to abbreviate mega, we can't really say that an uppercase M indicates binary megabytes unless we also say that a lowercase m indciates decimal megabytes. So the unit symbols for binary and decimal megabytes, gigabytes, etc. are the same. Bytes and bits can be distinguished by an uppercase B for bytes and a lowercase b for bits. This may not be strictly followed. Note that even the rules that are established are often broken, so you should be careful about assuming bits or bytes and decimal or binary based on how someone capitalizes these unit symbols. Much of this is not clearly defined and even that which is will often be stated incorrectly in references such as computer dictionaries. ---------------------------------------------------------------------- Data Transmission Bandwidth Measurement The bandwidth or data-carrying capacity of a channel such as a network cable, a modem, or a phone line is commonly measured in bits per second. kilobit (kb) = 1,000 bits Kilobit (Kb) = 1,024 bits bps = bits per second kbps = kilobits (1000 bits) per second (decimal) Kbps = Kilobits (1024 bits) per second (binary) kBps = kilobytes (1000 bytes) per second KBps = Kilobytes (1024 bytes) per second Mbps = megabits (1,000,000 bits) per second (decimal) Mbps = megabits (1,048,576 bits) per second (binary) MBps = megabytes (1,000,000 bytes) per second (decimal) MBps = megabytes (1,048,576 bytes) per second (binary) Here again, the lowercase k is used in its standard way to indicate decimal kilobits and the uppercase K is used to indicate binary kilobits. Decimal kilobits and megabits would normally be used to indicate bandwidth. As an example, the speed of a 14,400 bps modem could also be indicated in kbps as 14.4 kbps. This measurement is commonly misused. The abbreviation kbps is often incorrectly capitalized when indicating decimal kbps, such as in modem speeds. It is also common to find kbps incorrectly defined as 1024 bytes per second. There is no such thing as binary bps and decimal bps - the number of bits per second is the number of bits per second. When bps is expressed as kbps, such as 14.4 kbps instead of 14,400 bps, this is obviously a straight decimal conversion, accomplished by a move of the decimal point. Modem speeds are correctly indicated in bps rates and not in baud rates. Even though not technically correct, the term baud is sometimes used interchangeably with bps. Although not normally used to measure modem speed, it is possible to use binary Kilobits, Kilobytes, megabits, and megabytes when measuring bandwidth (data throughput). ---------------------------------------------------------------------- Frequency Measurement Frequency is measured in hertz. The term hertz replaces an older term, cps (cycles per second), which was used until the 1960's, when it was changed to honor Heinrich Hertz. The unit symbol Hz is capitalized, the word hertz is not. A hertz is defined as one cycle per second. In AC (alternating current) terms, a cycle is the completion of one sine wave. In digital electronics, which uses DC (direct current), a cycle is defined by a clock pulse generated by a crystal. In a PC, the speed of the CPU and the speed of various busses on the motherboard are measured in megahertz. For example, a 486DX-33 CPU chip is rated to operate at speeds of up to 33MHz. hertz (Hz) = 1 cycle per second kilohertz (kHz) = 1,000 hertz megahertz (MHz) = 1,000,000 hertz gigahertz (GHz) = 1,000,000,000 hertz There is a reciprocal relationship between megahertz and nanoseconds: 1000/MHz = ns 1000/ns = MHz Frequency in MHz Length of Clock Cycle 1 MHz 1000 ns 8 MHz 125 ns 16 MHz 62.5 ns 20 MHz 50 ns 25 MHz 40 ns 33 MHz 33 ns 100 MHz 10 ns 1000 MHz 1 ns ---------------------------------------------------------------------- Access Time Measurement millisecond (ms) = 1 thousandth of a second microsecond (us) = 1 millionth of a second nanosecond (ns) = 1 billionth of a second Memory chips have a speed rating to indicate access time. This is normally indicated in nanoseconds. EPROM chips typically have an access time of around 200 ns. DRAM chips, such as on a SIMM module, typically have an access time of around 70 ns. SRAM chips, such as used in external cache subsystems, typically have an access time of 10, 15, or 20 ns. Hard drives have an access time rating which is usually expressed in milliseconds. This ranges from 200 ms on older drives to 10 ms on newer drives. ---------------------------------------------------------------------- . # ====================================================================== (C) 1997-1998 Micro Firmware, Inc.