ASIC (Application Specific Integrated
Circuit)
A chip that is custom designed for a specific application rather than a
general-purpose chip such as a microprocessor. The use of ASICs improve
performance over general-purpose CPUs, because ASICs are "hardwired" to
do a specific job and do not incur the overhead of fetching and interpreting
stored instructions. An ASIC chip performs an electronic operation as fast
as it is possible to do so, providing, of course, that the circuit design
is efficiently architected.
ASSP (Application Specific Standard Part/Product)
Another term for an ASIC chip.
CSIC (Customer Specific Integrated Circuit)
Pronounced "C-sick." Another term for ASIC, which was
coined by Motorola. Some feel this is a more accurate description of an
ASIC chip, since ASICs can be used for a variety of purposes. CSIC is somewhat
a redundant reference to an ASIC chip. ASICs are already customized for
a specific use.
gate array
An unfinished chip with electronic components that have not been connected.
The chip is completed by designing and adhering the top metal layers which
provide the interconnecting pathways. This final masking stage is less
costly than designing the chip from scratch. The gate array is made up
of basic cells, each cell containing some number of transistors and resistors
depending on the vendor. Using a cell library (gates, registers, etc.)
and a macro library (more complex functions), the customer designs the
chip, and the vendor's software generates the masks that connect the transistors.
hard macro
The design of a logic function that specifies how the required logic elements
are interconnected and specifies the physical pathways and wiring patterns
between the components. Also called a "macro cell."
soft macro
The design of a logic function that specifies how the required logic elements
are interconnected, but not the physical wiring pattern on the chip. Contrast
with hard macro.
NVM (non-volatile memory)
Memory that holds its content without power. ROMs, PROMs,
EPROMs
and flash memory are examples. Disks and tapes may be called non-volatile
memory, but they are usually considered storage devices. Sometimes the
term refers to memory that is inherently volatile, but maintains its content
because it is connected to a battery at all times.
ROM (Read Only Memory)
A memory chip that permanently stores instructions and data. Its contents
are created at the time of manufacture and cannot be altered. ROM chips
are used to store control routines in personal computers (ROM BIOS), peripheral
controllers and other electronic equipment. They are also often the sole
contents inside a cartridge that plugs into printers, video games and other
systems. When computers are used in hand-held instruments, appliances,
automobiles and any other such devices, the instructions for their routines
are generally stored in ROM chips or some other non-volatile chip such
as a PROM or EPROM. Instructions
may also be stored in a ROM section within a general-purpose computer on
a chip.
PROM (Programmable Read Only Memory)
A permanent memory chip that is programmed, or filled, by the customer
rather than by the chip manufacturer. It differs from a ROM, which is programmed
at the time of manufacture. PROMs have been mostly superseded by EPROMs,
which can be reprogrammed.
EPROM (Erasable Programmable
ROM)
A programmable and reusable chip that holds its content until erased under
ultraviolet light. EPROMS have a lifespan of a few hundred write cycles.
EPROMS are expected to eventually give way to flash memory.
EEPROM (Electrically Erasable
Programmable Read Only Memory)
A memory chip that holds its content without power. It can be erased,
either within the computer or externally and usually requires more
voltage for erasure than the common +5 volts used in logic circuits. It
functions like non-volatile RAM, but writing to EEPROM is slower than writing
to RAM. EEPROMs are used in devices that must keep data up-to-date without
power. For example, a price list could be maintained in EEPROM chips in
a point of sale terminal that is turned off at night. When prices change,
the EEPROMs can be updated from a central computer during the day. EEPROMs
have a lifespan of between 10K and 100K write cycles.
flash memory
A memory chip that holds its content without power. Derived from the EEPROM
chip technology, which can be erased in place, flash memory is less
expensive and more dense. Unlike DRAM and SRAM memory chips, in which
a single byte can be written, flash memory must be erased and written in
fixed blocks, typically ranging from 512 bytes up to 256KB. The term was
coined by Toshiba for its ability to be erased "in a flash." Flash chips
have been used in various communications and industrial products as well
as to replace ROM BIOS chips so that the BIOS could be updated. Flash chips
generally have lifespans from 100K to 300K write cycles. Flash memory chips
are conveniently packaged as "flash cards," using the PC Card, CompactFlash,
Smart
Media and similar formats. They have become widely used as film in
digital cameras as well as auxiliary storage in a variety of handheld commercial
and consumer devices. There are two types of flash interfaces. The earlier
linear
flash, which is also used to execute a program directly from the chip (XIP),
requires Flash Translation Layer (FTL) or Flash File System (FFS) software
to make it look like a disk drive. There are various implementations of
FTL and FFS. The second type is the ATA interface, which is widely used
for hard disks and has the same 512-byte block (sector) size.
Smart Media
An ultra-compact flash memory format developed
by Toshiba. About the size of CompactFlash, but as thin as a credit card,
Smart Media cards are popular storage for digital cameras. Available in
3.3 and 5 volt variations, Smart Media cards require no assembly in manufacture
as they are actually flash memory chips in a unique chip package. The cards
can be plugged into a Smart Media socket or into a standard Type II PC
Card slot with an adapter. Capacities up to 64MB are expected by late 1999.
Memory Stick
A flash memory card from Sony designed for handheld
digital appliances such as cameras and camcorders. Introduced in 1998 with
4 and 8MB capacities, the tiny modules are less than 1x2" and about a tenth
of an inch thick (.85 x 1.97 x .11"). Transfer to a PC is made via a PC
Card adapter.
PROM programmer
A device that writes instructions and data into PROM
chips. The bits in a new PROM are all 1s (continuous lines). The PROM programmer
only creates 0s, by "blowing" the middle out of the 1s. Some earlier units
were capable of programming both PROMs and EPROMs.
FPGA (Field Programmable Gate Array)
A programmable logic chip (PLD) with a high density
of gates. There are a variety of FPGA architectures, some of which can
be very sophisticated, including not only programmable logic blocks, but
programmable interconnects and switches between the blocks.
PLD (Programmable Logic Device)
A logic chip that is programmed at the customer's site. It uses either
fusible links, which are "blown" to open the lines or Actel's Antifuse
technology to fuse them together. There are a wide variety of PLD techniques;
however, most PLDs are of the PLA (Programmable Logic Array) or PAL (Programmable
Array Logic) variety, which provide different configurations of AND and
OR gates. Unlike gate arrays, which require the
final masking fabrication process, PLDs are easily programmable in the
field. PLDs are always used for logical functions, but programmable storage
chips such as PROMs and EPROMs
may also be considered as PLDs if they contain program code rather than
just data.
SPLD (Simple PLD)
A programmable logic device that provides an array of logic blocks that
can be programmed.
CPLD (Complex PLD)
A programmable logic device that includes a programmable interconnect between
the logic blocks.
Antifuse
A PLD technology from Actel
Corporation, that works the opposite of typical programmable chip methods.
Instead of creating open circuits (blowing the fuse), closed circuits are
created. Two metal layers sandwich a layer of non-conductive, amorphous
silicon. When voltage is applied to this middle layer, the amorphous silicon
is turned into polysilicon, which is conductive.
EDA (Electronic Design Automation)
Using the computer to design and simulate the performance of electronic
circuits on a chip.
VHDL (VHSIC Hardware Description Language)
A hardware description language (HDL) used to design electronic systems
at the component, board and system level. VHDL allows models to be developed
at a very high level of abstraction. Initially conceived as a documentation
language only, most of the language can today be used for simulation and
logic synthesis. VHDL is an IEEE standard,
but was initially developed for the U.S. military's VHSIC
program.
Verilog
A hardware description language (HDL) used to design electronic systems
at the component, board and system level. Developed by Phil Morby, it was
initially developed as a simulation language that could also be used to
describe the stimulus (test factors). By the late 1980s, Verilog was the
de facto standard for proprietary HDLs. After Cadence
put it into the public domain, it became an IEEE standard. A Verilog simulator
is a separate entity, but the term by itself typically refers to the language
(HDL).
VHSIC (Very High Speed Integrated
Circuit)
Pronounced "vizik." Ultra-high-speed chips employing LSI and VLSI technologies.
logic synthesis
The conversion of a high-level electronic circuit description into a list
of logic gates and their interconnections, called the netlist.
Every logic synthesis program understands some subset of Verilog
and VHDL. The market leader in logic synthesis software
is Synopsis.
silicon compiler
Software that translates the electronic design of a chip into the layout
of the logic gates, including the actual masking from one transistor to
another. The source of the compilation is either a high-level description
or the netlist.
netlist
A list of logic gate and their interconnections which make up a circuit.
back end
The support components of a computer system. It typically refers to the
database management system (DBMS), which is the storehouse for the data.
front end
The head, starting point or input side in a system. For example, it may
refer to the graphical interface on a user's workstation where all data
is entered or to a communications system, such as a front end processor
or TP monitor that accepts incoming transactions and messages.
MCM (MultiChip Module or MicroChip Module)
A chip housing that uses a ceramic base and contains two or more raw chips
closely connected with high-density lines. This packaging method saves
space and speeds processing due to short leads between chips. MCMs were
originally called microcircuits or hybrid
microcircuits, since this technique was suited for mixing analog and
digital components together. MCMs offer a more workable solution to wafer
scale integration, in essence, building the "superchip," which has been
very difficult to implement.
hybrid microcircuit
An electronic circuit composed of different types of integrated circuits
and discrete components, mounted on a ceramic base. Used in military and
communications applications, it is especially suited for building custom
analog circuits including A/D and D/A converters, amplifiers and modulators.
embedded system
An embedded system implies a fixed set of functions programmed into a non-volatile
memory (ROM, flash memory, etc.) in contrast to a general-purpose computing
machine.
realtime system
A computer system that responds to input signals fast enough to keep an
operation moving at its required speed.
RTOS (RealTime Operating System)
An operating system designed for use in a realtime computer system.
SONET (Synchronous Optical NETwork)
A fiber-optic transmission system for high-speed digital traffic. Employed
by telephone companies and common carriers, SONET speeds range from 51
megabits to multiple gigabits per second. SONET is an intelligent system
that provides advanced network management and a standard optical interface.
It uses a self-healing ring architecture that is able to reroute traffic
if a line goes down. SONET backbones are widely used to aggregate lower-speed
T1
and T3 lines. SONET is specified in the Broadband ISDN
(BISDN) standard. The European counterpart is SDH.
Following are the levels of service. OC (Optical Carrier) refers to the
optical signal, and STS (Synchronous Transport Signal) refers to the electrical
signal, which is the same speed.
SONET Circuits |
Service Speed (Mbps) |
Implementation |
OC-1, STS-1 |
51.84 |
28 DS1s or 1 DS3 |
OC-3, STS-3 |
155.52 |
3 STS-1s |
OC-3c, STS-3c |
155.52 |
Concatenated |
OC-12, STS-12 |
622.08 |
12 STS-1, 4 STS-3 |
OC-12c, STS-12c |
622.08 |
12 STS-1, 4 STS-3c |
OC-48, STS-48 |
2488.32 |
48 STS-1, 16 STS-3 |
OC-192, STS-192 |
9953.28 |
192 STS-1, 64 STS-3 |
OC-768, STS-768 |
38813.12 |
768 STS-1, 256 STS-3 |
SDH (Synchronous Digital Hierarchy)
The European counterpart to SONET.
ISDN (Integrated Services Digital
Network)
An international telecommunications standard for transmitting voice, video
and data over digital lines running at 64 Kbps. The telephone companies
commonly use a 64 Kbps channel for digitized, two-way voice conversations.
ISDN service is available in most parts of the U.S. ISDN uses 64 Kbps circuit-switched
channels, called B channels, or "bearer" channels, to carry voice and data.
It uses a separate D channel, or "delta," channel for control signals.
The D channel is used to signal the telephone company computer to
make calls, put them on hold and activate features such as conference calling
and call forwarding. It also receives information about incoming calls,
such as the identity of the caller. ISDN's basic service is Basic Rate
Interface (BRI), which is made up of two 64 Kbps B channels and one 16
Kbps D channel (2B+D). If both channels are combined into one, called bonding,
the total data rate becomes 128 Kbps and is four and a half times the bandwidth
of a V.34 modem (28.8 Kbps). ISDN's high-speed service is Primary Rate
Interface (PRI). It provides 23 B channels and one 64 Kbps D channel (23B+D),
which is equivalent to the 24 channels of a T1 line.
When several channels are bonded together, high data rates can be achieved.
For example, it is common to bond six channels for quality videoconferencing
at 384 Kbps. In Europe, PRI includes 30 B channels and one D channel, equivalent
to an E1 line.
BISDN (Broadband ISDN)
A second-generation ISDN standard that uses fiber-optic
cables for speeds of 155 Mbps and higher. BISDN's bottom three layers of
implementation comprise ATM (asynchronous transfer mode),
which by itself is gaining ground as a networking technology for LANs and
WANs.
CO (central office)
A local telephone company switching facility that covers a geographic area
such as a small town or a part of a city. It is where subscribers' telephone
lines in the local loop are connected to intracity and intercity trunks.
There are more than 25,000 central offices in the U.S.
PSTN (Public Switched Telephone Network)
The worldwide voice telephone network. Once only an analog system, the
heart of most telephone networks today is all digital. In the U.S., most
of the remaining analog lines are the ones from your house or office to
the telephone company's central office (CO).
AMPS (Advanced Mobile Phone Service)
The analog cellular mobile phone system in North and South America and
more than 35 other countries. It uses the FDMA transmission technology.
AMPS is the cellular equivalent of POTS.
POTS (Plain Old Telephone Service)
VPN (Virtual Private Network)
A private network that is configured within a public network. For years,
common carriers have built VPNs that appear as private national or international
networks to the customer, but physically share backbone trunks with other
customers. VPNs have been built over X.25, Switched
56, frame relay and ATM
technologies. Today, there is tremendous interest in building VPNs over
the Internet. In order to maintain privacy in a public environment, VPNs
use access control and encryption.
broadband
(1) High-speed transmission. The term is commonly used to refer to communications
lines or services at T1 rates (1.544 Mbps) and above.
(2) A technique for transmitting data, voice and video using the same frequency
division multiplexing (FDM) technique as cable TV. Modems are required
for this method, because the digital data has to be modulated onto the
line.
baseband
A communications technique in which digital signals are placed onto the
transmission line without change in modulation. It is usually limited
to a few miles and does not require the complex modems used in broadband
transmission. Common baseband LAN techniques are token passing ring (Token
Ring) and CSMA/CD (Ethernet). In baseband, the full
bandwidth of the channel is used, and simultaneous transmission of multiple
sets of data is accomplished by interleaving pulses using TDM (time division
multiplexing). Contrast with broadband transmission, which transmits data,
voice and video simultaneously by modulating each signal onto a different
frequency, using FDM (frequency division multiplexing).
narrowband
In communications, transmission rates from 50 bps to 64 Kbps. Earlier uses
of the term referred to 2,400 bps or less or to sub-voice grade transmission
from 50 to 150 bps. Contrast with wideband and broadband.
wideband
In communications, transmission rates from 64 Kbps to 2 Mbps.
SIMD (Single Instruction stream
Multiple Data stream)
A computer architecture that performs one operation on multiple sets of
data, for example, an array processor. One computer or processor is used
for the control logic and the remaining processors are used as slaves,
each executing the same instruction. Contrast with MIMD.
MIMD (Multiple Instruction stream
Multiple Data stream)
A computer architecture that uses multiple processors, each processing
its own set of instructions simultaneously and independently of the others.
Contrast with SIMD.
VLIW (Very Long Instruction Word)
A CPU architecture that reads a group of instructions and executes them
at the same time. For example, the group (word) might contain four instructions,
and the compiler ensures that those four instructions are not dependent
on each other so they can be executed simultaneously. Otherwise, it places
no-ops (blank instructions) in the word where necessary.
prediction
A VLIW concept. In CPU instruction execution, predicting
the outcome of a branch so that those instructions may be executed in parallel
with the current instructions. If the CPU guesses the wrong branch, it
will take extra machine cycles to go back and execute the correct one;
however, on average, if the prediction algorithms are good, overall performance
is increased.
EPIC (Explicitly Parallel Instruction
Computing)
The parallel architecture used in Intel's IA-64 chips.
It was originally developed by HP.
IA-64 (Intel Architecture-64)
The 64-bit architecture used in Intel's upcoming Merced
chip. Instead of the variable-length instructions of the x86, which are
from one to 14 bytes in length, IA-64 uses four-byte, fixed-length instructions
bundled in sets of three (long instruction words). IA-64 chips use 256
registers for integer and floating point operations compared to 16 in the
x86. It also employs a technique called predication,
where both sides of a branch instruction are executed in parallel. When
the correct branch is determined, the results for the incorrect side are
discarded. With IA-64, compilers have to be much more intelligent, placing
codes into the instruction bundles that tell the CPU how to execute instructions
in parallel. They also have to place instructions in interleaved order
for predication. If predication is not set up, the CPU will perform traditional
branch prediction, whereby it attempts to guess the outcome of a branch
and executes those instructions in parallel. IA-64 also supports speculative
loading, which loads data into its registers before the instructions actually
need to process it. Explicitly Parallel Instruction Computing (EPIC)
is Intel's brand name for its new parallel architecture.
Merced
Intel's next-generation CPU architecture, also known as the P7. Expected
in 2000, it is jointly designed by Intel and HP. Although it uses a new
64-bit instruction architecture (IA-64), Merced is
also expected to run x86 and PA-RISC software natively with clock speeds
at 600MHz and beyond. In order to take full advantage of the 64-bit architecture,
applications have to be upgraded. Merced chips should contain more than
10 million transistors using .18 micron technology, and its IA-64 architecture
is designed for fast parallel instruction execution. The bottom line is
that Merced was designed at the end of the 20th century, whereas x86 chips
hark back to an architecture from the early 1970s, which was based on a
fraction of the number of transistors available today. The Merced should
be a formidable product. In addition, not only will HP-UX and Windows NT
run on Merced, but Digital Unix, SCO UnixWare and Solaris are expected
to be ported to the new chip family.
superscalar
A CPU architecture that allows more than one instruction to be executed
in one clock cycle.
DSP (Digital Signal Processor)
A special-purpose CPU used for digital signal processing. It provides ultra-fast
instruction sequences, such as shift and add, and multiply and add, which
are commonly used in math-intensive signal processing applications. DSP
chips are widely used in myriads of devices, including sound cards, fax
machines, modems, cellular phones, high-capacity hard disks and digital
TVs (see definition 2 below). The first DSP chip used in a commercial product
was believed to be from TI, which was
used in its very popular Speak & Spell game in the late 1970s.
firmware
A category of memory chips that hold their content without electrical power
and include ROM, PROM, EPROM and
EEPROM
technologies. Firmware becomes "hard software" when holding program code.
yacc (Yet Another Compiler Compiler)
A UNIX compiler that is used to create C compilers. Part of its code is
included in the generated compiler.
bison
The Free Software Foundation's version of yacc.
DSLAM (DSL Access Multiplexor)
A central office (CO) device for ADSL
service that intermixes voice traffic and DSL traffic
onto a customer's DSL line. It also separates incoming phone and data signals
and directs them onto the appropriate carrier's network.
DS (Digital Signal)
A classification of digital circuits. The DS technically refers to the
rate and format of the signal, while the T designation refers to the equipment
providing the signals. In practice, "DS" and "T" are used synonymously;
for example, DS1 and T1.
NORTH AMERICA, JAPAN, KOREA, ETC.
Voice |
Service Channels |
Speed |
DS0 |
1 |
64 Kbps |
DS1 |
24 |
1.544 Mbps (T1) |
DS1C |
48 |
3.152 Mbps (T1C) |
DS2 |
96 |
6.312 Mbps (T2) |
DS3 |
672 |
44.736 Mbps (T3) |
DS4 |
4032 |
274.176 Mbps (T4) |
EUROPE (ITU)
Voice |
Service Channels |
Speed (Mbps) |
E1 |
30 |
2.048 |
E2 |
120 |
8.448 |
E3 |
480 |
34.368 |
E4 |
1920 |
139.264 |
E5 |
7680 |
565.148 |
set-top box
The cable TV box that "sits on top" of the TV set. A variety of new set-top
boxes are emerging for Internet TV and other interactive services.
MMU (Memory Management Unit)
A virtual memory circuit that translates logical addresses into physical
addresses.
NIS (Network Information Services)
A naming service from SunSoft that allows resources to be easily added,
deleted or relocated. Formerly called Yellow Pages, NIS is a de facto UNIX
standard. NIS+ is a redesigned NIS for Solaris 2.0 products. The combination
of TCP/IP, NFS and NIS comprise the primary networking components of UNIX.
NFS (Network File System)
The UNIX networking protocol that allows files and printers to be shared
across the network. This de facto UNIX standard, which is widely known
as a distributed file system, was developed by Sun. NFS services are also
available on Windows NT servers, which enables UNIX workstations to gain
access to its files and printers. The counterpart of NFS in the DOS, Windows
and OS/2 world is SMB (Server Message Block), and in NetWare, NCP (NetWare
Core Protocol). NFS, SMB and NCP are all high-level protocols (layers 5
to 7) that provide open and close file and read and write functions as
well as access control. They ride on top of the transport protocols (TCP/IP,
NetBEUI, IPX/SPX), which manage the transfer. See WebNFS, SMB and NCP.
DSU/CSU (Digital Service Unit/Channel
Service Unit)
A pair of communications devices that connect an inhouse line to an external
digital circuit (T1, DDS, etc.). It
is similar to a modem, but connects a digital circuit rather than an analog
one. The CSU terminates the external line at the customer's premises. It
also provides diagnostics and allows for remote testing. If the customer's
communications devices are T1 ready and have the proper interface, then
the CSU is not required, only the DSU. The DSU does the actual transmission
and receiving of the signal and provides buffering and flow control. The
DSU and CSU are often in the same unit. The DSU may also be built into
the multiplexor, commonly used to combine digital signals for high-speed
lines.
Switched 56
A dial-up digital service provided by local and long distance telephone
companies. There is a monthly fee and per-minute charge like the analog
voice network. For connection, a DSU/CSU is used
instead of a modem. Switched 56 uses a 64 Kbps channel, but one bit per
byte is used for in band signaling, leaving 56 Kbps for data.
X.25
An ITU standard (1976) for packet switching networks. Public X.25
communications networks have been available worldwide for many years, which
provide a switched data service at 56 Kbps or less. Such networks are widely
used for point of sale (POS) terminals, credit card verifications and automatic
teller machine transactions. New packet-switched networks employ frame
relay and SMDS technologies rather than X.25.
frame relay
A high-speed packet switching protocol used in wide area networks
(WANs). It has become popular for LAN to LAN connections across remote
distances, and services are provided by all the major carriers. Frame relay
is faster than traditional X.25 networks, because
it was designed for today's reliable circuits and performs less rigorous
error detection. Frame relay provides for a granular service up to DS3
rates of 44.736 Mbps and is suited for data and image transfer. Because
of its variable-length packet architecture, it is not the most efficient
technology for realtime voice and video.
SMDS (Switched Multimegabit Data
Service)
A high-speed, switched data communications service offered by the local
telephone companies for interconnecting LANs in different locations. It
was introduced in 1992 and became generally available nationwide by 1995.
Connection to an SMDS service can be made from a variety of devices, including
bridges, routers, CSU/DSUs as well as via frame relay
and ATM networks. SMDS can employ various networking
technologies. Early implementations use the IEEE 802.6 DQDB MAN technology
at rates up to 45 Mbps. Data is framed for transmission using the SMDS
Interface Protocol (SIP), which packages data as Level 3 Protocol Data
Units (L3_PDU). The L3_PDU contains source and destination addresses and
a data field that holds up to 9188 bytes.
ATM (Asynchronous Transfer Mode)
A network technology for both LANs and WANs that supports realtime voice
and video as well as data. The topology uses switches that establish a
logical circuit from end to end, which guarantees a quality of service
(QoS) for that transmission. However, unlike telephone
switches that dedicate circuits end to end, unused bandwidth in ATM's logical
circuits can be appropriated whenever available. For example, idle bandwidth
in a videoconference circuit can be used to transfer data. ATM is also
highly scalable and supports transmission speeds of 1.5, 25, 100, 155,
622 and 2488 Mbps. ATM is also running as slow as 9.6 Kbps between ships
at sea. An ATM switch can be added into the middle of a switch fabric to
enhance total capacity, and the new switch is automatically updated using
ATM's PNNI routing protocol.
Cell Switching
ATM works by transmitting all traffic as fixed-length, 53-byte cells.
This fixed unit allows very fast switches to be built, because the processing
associated with variable-length packets is eliminated (finding the end
of the frame). The small ATM packet also ensures that voice and video can
be inserted into the stream often enough for realtime transmission.
Quality of Service (Qos)
The ability to speficy a quality of service is one of ATM's most important
features, allowing voice and video to be transmitted smoothly. Constant
Bit Rate (CBR) guarantees bandwidth for realtime voice and video. Available
Bit Rate (ABR) adjusts bandwidth according to congestion levels for LAN
traffic. Unspecified Bit Rate (UBR) provides a best effort for remote users.
Realtime variable Bit Rate (rt-VBR) supports interactive multimedia that
requires minimal delays, and non-realtime variable bit rate (nrt-VBR) is
used for bursty transaction traffic.
Integration with Legacy LANs
Network applications use protocols, such as TCP/IP, IPX, AppleTalk
and DECnet, and there are tens of millions of Ethernet
and Token Ring clients in existence. ATM has to coexist with these legacy
protocols and networks. MPOA is an ATM standard that
routes legacy protocols while preserving ATM quality of service. LANE (LAN
Emulation) interconnects legacy LANs by encapsulating Ethernet and Token
Ring frames into LANE packets and then converting them into ATM cells.
It supports existing protocols without changes to Ethernet and Token Ring
clients, but uses MPOA route servers or traditional routers for internetworking
between LAN segments.
MPOA (MultiProtocol Over ATM)
An ATM Forum standard that provides routing of legacy
protocols (IP, IPX, etc.) over ATM networks. MPOA separates the routing
processing from the actual forwarding. A route server performs the routing
calculations and sends its results to the ATM switches and edge devices
which perform high-speed forwarding of the packets.
PNNI (Private Network-to-Network
Interface)
A routing protocol used between ATM switches in an ATM
network. It lets the switches inform each other about network topology
so they can make appropriate forwarding decisions. PNNI is based on the
OSPF
protocol, but also measures line capacities and delays rather than just
simple cost metrics. Thus, ATM switches can dynamically reroute packets
based on current line conditions.
OSPF (Open Shortest Path First)
A routing protocol that determines the best path for routing IP traffic
over a TCP/IP network. OSPF is an interior gateway protocol (IGP),
which is designed to work within an autonomous system. It is also a link
state protocol that provides less router to router update traffic than
the RIP protocol (distance vector protocol) that it
was designed to replace.
RIP (Routing Information Protocol)
A simple routing protocol that is part of the TCP/IP protocol suite. It
determines a route based on the smallest hop count between source and destination.
RIP is a distance vector protocol that routinely broadcasts routing information
to its neighboring routers and is known to waste bandwidth.
AppleTalk,
DECnet,
TCP/IP, NetWare and VINES all use incompatible versions of RIP.
IGP (Interior Gateway Protocol)
A broad category of routing protocols that support a single, confined geographic
area such as a local area network (LAN). Contrast with EGP.
EGP (Exterior Gateway Protocol)
A broad category of routing protocols that are designed to span different
autonomous systems. Contrast with IGP.
IGRP (Interior Gateway Routing
Protocol)
A proprietary routing protocol from Cisco
that was developed in 1988 to overcome the shortcomings of RIP.
IGRP takes bandwidth, latency, reliability and current traffic load into
consideration. It is typically used within an autonomous system, such as
an Internet domain. IGRP was superseded by Enhanced IGRP (EIGRP), which
provides enhancements such as the ability to detect a loop in the network.
AppleTalk
Apple's local area network architecture introduced in 1985. It supports
Apple's proprietary LocalTalk access method as
well as Ethernet and Token Ring. The AppleTalk network manager and the
LocalTalk access method are built into all Macintoshes and LaserWriters.
With other products from Apple and third parties, AppleTalk can run in
PCs, VAXs and UNIX workstations. Since AppleTalk is patterned after the
OSI model, it is a routable protocol that contains a network layer (OSI
layer 3).
DECnet
Digital's communications network, which supports Ethernet-style LANs and
baseband and broadband WANs over private and public lines. It interconnects
PDPs,
VAXs,
PCs, Macs and workstations. In DECnet philosophy, a node must be an intelligent
machine and not simply a terminal as in other systems. DECnet/DOS allows
DOS machines to function as end nodes in DECnet networks, and DECnet/OSI
is the implementation of DECnet Phase V that supports OSI and provides
compatibility with DECnet Phase IV and TCP/IP.
LocalTalk
A LAN access method from Apple that uses twisted pair wires and transmits
at 230,400 bps. It runs under AppleTalk and uses
a daisy chain topology that can connect up to 32 devices within a distance
of 1,000 feet. Third party products allow it to hook up with bus, passive
star and active star topologies. Apple's LocalTalk PC Card lets a PC gain
access to an AppleTalk network.
VINES (VIrtual NEtworking System)
A UNIX System V-based network operating system from Banyan
Systems Inc., Westboro, MA, that runs on DOS and OS/2-based servers.
It provides internetworking of PCs, minis, mainframes and other computer
resources providing information sharing across organizations of unlimited
size. Incorporating mainframe-like security with a global directory service
called Streettalk, VINES allows access to all network users and resources.
Options include printer sharing, e-mail, remote PC dial-in, bridges and
gateways.
VAX (Virtual Address eXtension)
A family of 32-bit computers from Digital
introduced in 1977 with the VAX-11/780 model. VAXes range from desktop
personal computers to mainframes all running the same VMS operating system.
Large models can be clustered in a multiprocessing environment to serve
thousands of users. Software compatibility between models caused the VAX
family to achieve outstanding success during the 1980s. VAXes also provide
PDP
emulation.
PDP (Programmed Data Processor)
A minicomputer family from Digital
that started with the 18-bit PDP-1 in 1959. Its $120,000 price was much
less than the million dollar machines of the time and 50 units were built.
In 1965, Digital legitimized the minicomputer industry with the PDP-8,
which sold for about $20,000. By the late 1970s, the PDP-8 processor was
put on a single chip and used in DECmate workstations. Other PDPs were
built, including 12-bit, 18-bit and 36-bit machines, the larger ones evolving
into DECsystem models. In 1970, Digital introduced the 16-bit PDP-11, which
became the most widely used minicomputer with more than 50,000 systems
sold. The VAX series was introduced in 1977, but PDP
machines lingered on for many years.
PC Card
A credit-card sized, removable module for portable computers standardized
by PCMCIA. PC Cards are also known as "PCMCIA cards."
PC Cards are 16-bit devices that are used to attach modems, network adapters,
sound cards, radio transceivers, solid state disks and hard disks to a
portable computer. The PC Card is a "plug and play" device, which is configured
automatically by the Card Services software. All PC Cards are 85.6 mm long
by 54 mm wide (3.37" x 2.126") and use a 68-pin connector. The original
Type I card is 3.3 mm thick and is typically used to hold memory. Type
II cards (5.0 mm thick) are commonly used for memory, modems and LAN adapters
in laptops. Type III cards (10.5 mm thick) are used to hold a hard disk,
wireless transceiver or other peripheral that needs more space. The Type
III slot can hold two Type II cards. Toshiba introduced a 16 mm Type IV
card, but this has not been officially adopted by the PCMCIA. Smaller cards
will work in a Type IV slot.
Card and Socket Services
In order to use a PC Card slot in the computer, Card and Socket services
must be loaded, typically at system startup. Card and Socket Services software
is generally included with laptops that have PC Card slots. It also comes
packaged with PC Cards. Card Services manage system resources required
by the PC Card, and, on PCs, determines which IRQs and memory and I/O addresses
are assigned. They also manage hot swapping and pass changes in events
to higher-level drivers written for specific cards. Card Services talk
to Socket Services, which is the lowest level of software that communicates
directly with the PC Card controller chips. Socket Services can be built
into the system BIOS or added via software.
CardBus
In early 1995, PCMCIA introduced the 32-bit CardBus standard. Although
electrically different, the CardBus is architecturally identical to the
PCI bus. The CardBus supports bus mastering and accommodates cards operating
at different voltages. Its advanced power managment features allows the
computer to take advantage of CardBus cards designed to idle or turn off
in order to increase battery life. The CardBus specification allows data
transfer up to 132 Mbytes/sec over a 33MHz, 32-bit data path.
PCMCIA (Personal Computer Memory
Card International Association)
An international standards body and trade association that was founded
in 1989 to establish a standard for connecting peripherals to portable
computers. PCMCIA created the PC Card. For more information,
visit www.pc-card.com.
DDS (Dataphone Digital Service)
An AT&T private line digital service with data rates from 2400 bps
to 56Kbps. Private analog lines can be connected to DDS lines.
T1
A 1.544 Mbps point-to-point dedicated, digital circuit provided
by the telephone companies. The monthly cost is typically based on distance.
T1 lines are widely used for private networks as well as interconnections
between an organization's PBX or LAN and the telco.
The first T1 line was tariffed by AT&T in January 1983. However, starting
in the early 1960s, T1 was deployed in intercity trunks by AT&T to
improve signal quality and make more efficient use of the network. A T1
line uses two wire pairs (one for transmit, one for receive) and
time division multiplexing (TDM) to interleave 24 64-Kbps voice or data
channels. The standard T1 frame is 193 bits long, which holds 24 8-bit
voice samples and one synchronization bit with 8,000 frames transmitted
per second. T1 is not restricted to digital voice or to 64 Kbps data streams.
Channels may be combined and the total 1.544 Mbps capacity can be broken
up as required.
PBX (Private Branch eXchange)
An inhouse telephone switching system that interconnects telephone
extensions to each other, as well as to the outside telephone network.
It may include functions such as least cost routing for outside calls,
call forwarding, conference calling and call accounting. Modern PBXs use
all-digital methods for switching and can often handle digital terminals
and telephones along with analog telephones.
CSMA/CD (Carrier Sense Multiple
Access/Collision Detection)
The LAN access method used in Ethernet. When a device wants to gain access
to the network, it checks to see if the network is free. If it is not,
it waits a random amount of time before retrying. If the network is free
and two devices access the line at exactly the same time, their signals
collide. When the collision is detected, they both back off and each wait
a random amount of time before retrying.
CLEC (Competitive Local Exchange
Carrier)
An organization offering local telephone services. Although most
CLECs are established as a telecommunications service organization, any
large company, university or city government has the option of becoming
a CLEC and supplying its own staff with dial tone at reduced cost. It must
have a telephone switch, satisfy state regulations, pay significant filing
fees and also make its services available to outside customers. This was
all sanctioned by the Telecommunications Act of 1996. Contrast with
ILEC.
ILEC (Incumbent Local Exchange
Carrier)
A traditional local telephone company such as one of the Regional Bell
companies (RBOCs). Contrast with CLEC.
ELEC (Enterprise LEC)
An organization that is large enough (about 2500 or more employees) to
file for CLEC status and become its own customer. As
a CLEC, it can purchase telephone service at wholesale rates that it can
sell to itself as well as to others to further reduce costs. The Yankee
Group coined the term.
LEC (Local Exchange Carrier)
An organization that provides local telephone service, which includes the
RBOCs, large companies such as GTE and hundreds of small, rural telephone
companies. A LEC controls the service from its central office (CO)
to subscribers within a local geographic area.
DSL (Digital Subscriber Line)
A technology that dramatically increases the digital capacity of ordinary
telephone lines (the local loops) into the home or office. DSL speeds
are very much tied to the distance between the customer and the telco central
office. The technology is geared to Internet access with its asymmetric
versions (faster downstream than upstream) and short haul connections with
symmetric versions (same rate coming and going). Unlike ISDN,
which is also digital but travels through the switched telephone network,
DSL provides "always-on" operation. At the telco central office, DSL traffic
is aggregated in a unit called the DSL Access Multiplexor (DSLAM)
and forwarded to the appropriate ISP or data network. Although DSL technologies
are just emerging, there have been more versions and alphabet soup than
most any other new transmission technology. The widely-touted Asymmetrical
DSL (ADSL) is available in two modulation schemes: Discrete
Multitone (DMT) or Carrierless Amplitude Phase (CAP). DMT is expected
to prevail. ADSL and its variants can share the same line as the telephone,
because they use higher frequencies than the voice band. However, a POTSsplitter
must be installed on the customer's premises to separate the line between
voice and ADSL. A version of ADSL, known as G.lite or Universal
ADSL (also ADSL Lite and splitterless ADSL), eliminates the splitter,
but other phones on the line may have to plug into low-pass filters to
isolate them from the ADSL frequencies. Stay tuned! CDSL (Consumer DSL)
is an asymmetric flavor that supports regular 56 Kbps V.90
modem access if ADSL is not available. RADSL (Rate adaptive DSL) adjusts
speeds based on signal quality. IDSL (ISDN DSL) merely provides ISDN speeds.
It does not support voice or go through the switched telephone network
as does ISDN. HDSL (High Bit Rate DSL) has been the most widely-used
DSL technology, providing T1 speeds over existing twisted
pair without requiring the additional provisioning required for setting
up T1 circuits. SDSL (Single Line DSL) is an HDSL variation that uses only
one pair of cables instead of two. HDSL and SDSL cannot share lines
with telephones. VDSL (Very High Bit Rate DSL) is expected to be used as
the final drop from a fiber optic junction point to nearby customers. VDSL's
extremely high capacity would let an entire apartment or office complex
obtain high-bandwidth services using existing copper wires without having
to replace the entire infrastructure with optical fiber. Like ADSL, VDSL
can share the line with the telephone.
ASYMMETRIC DSL (Can share telephone line)
Type |
Upstream |
Downstream |
Pairs |
Cable Distance (ft) |
ADSL |
64-800 Kbps |
0.5-8 Mbps |
1 |
12-18000 |
ADSL Lite |
64-384 Kbps |
1-1.5 Mbps |
1 |
22-25000 |
CDSL |
128 Kbps |
1 Mbps |
1 |
12-18000 |
RADSL |
128-1024 Kbps |
0.6-7 Mbps |
1 |
18-25000 |
VDSL |
1.6 Mbps |
13 Mbps |
1 |
1000-4500 |
|
3.2 Mbps |
26 Mbps |
1 |
1000-4500 |
|
6.4 Mbps |
52 Mbps |
1 |
1000-4500 |
SYMMETRIC DSL (Cannot share telephone line except for VDSL)
Type |
Upstream & Downstream |
Pairs |
Cable Distance (ft) |
HDSL |
768 Kbps |
2 |
12000 |
|
1.544 Mbps (T1) |
2 |
12000 |
|
2.048 Mbps (E1) |
3 |
12000 |
HDSL-2 |
1.544 Mbps (T1) |
1 |
12000 |
|
2.048 Mbps (E1) |
1 |
12000 |
SDSL |
384 Kbps |
1 |
5-10000 |
|
768 Kbps |
1 |
5-10000 |
IDSL |
144 Kbps |
1 |
18000 |
VDSL** |
26 Mbps |
1 |
1000-4500 |
**can share telephone line
ATU (ADSL Transceiver Unit)
A device that provides ADSL (see DSL)
modulation of the telephone line. The device at the telco side is the ATU-C
(Central), which is a line card plugged into the DSLAM.
The unit at the customer's side is the ATU-R (Remote), which is either
an external modem or a card plugged into the PC.
V.90
An ITU standard (1998) for a modem that communicates at 56 Kbps downstream
and 33.6 Kbps upstream. It is intended for use only with ISPs and online
services that are digitally attached to the telephone system. Most service
providers are typically connected with high-speed digital T1
or T3 circuits. In practice, the downstream link isn't
generally faster than 45 Kbps in these PCM modems, so called because they
use pulse code modulation downstream and standard V.34 upstream. Initially,
two incompatible technologies competed in this arena: x2 from U.S. Robotics
and K56Flex from Rockwell and Lucent.
Such modems can be upgraded to V.90 if they contain software-upgradable
memory chips.
MAPI (Mail API)
A programming interface that enables an application to send and receive
mail over the Microsoft Mail messaging system. Simple MAPI is a subset
of MAPI that includes a dozen functions for sending and retrieving mail.
TAPI (Telephony API)
A programming interface from Microsoft and Intel that is part of Microsoft's
WOSA
architecture. It allows Windows client applications to access voice services
on a server. TAPI is designed to provide interoperability between PCs and
telephone equipment, including phone systems and PBXs.
WOSA (Windows Open System Architecture)
An umbrella term for a variety of programming interfaces from Microsoft
designed to provide application interoperability across the Windows environment.
It provides a common denominator for front-end Windows applications to
access back-end services from different vendors. For example, any WOSA-compliant
query program from one vendor can gain access to any WOSA-compliant DBMS
from any another vendor. See SPI. WOSA Interface Provides
access to: ODBC Databases (DBMSs), MAPI Messaging systems, TAPI
Telephone network services, LSAPI Software licensing, Windows SNA IBM SNA
networks, Windows Sockets Internet, TCP/IP networks, Microsoft RPC Run
remote procedures, Financial Services Banking services, WOSA/XRT News,
stock market, etc.
SPI (Service Provider Interface)
The programming interface for developing Windows drivers under WOSA.
In order to provide common access to services, the application (query,
word processor, e-mail program, etc.) is written to a particular WOSA-supported
interface, such as ODBC or MAPI, and the developer
of the service software (database manager, document manager, print spooler,
etc.) writes to the SPI for that class of service.
NetBIOS
The native networking protocol in DOS and Windows networks. Although originally
combined with its transport layer protocol (NetBEUI), NetBIOS today provides
a programming interface for applications at the session layer (layer 5).
NetBIOS can ride over NetBEUI, its native transport, which is not routable,
or over TCP/IP and SPX/IPX, which are routable protocols. NetBIOS computers
are identified by a unique 15-character name, and Windows machines (NetBIOS
machines) periodically broadcast their names over the network so that Network
Neighborhood can catalog them. For TCP/IP networks, NetBIOS names are turned
into IP addresses via manual configuration in an LMHOSTS file or a WINS
server. There are two NetBIOS modes. The Datagram mode is the fastest mode,
but does not guarantee delivery. It uses a self-contained packet with send
and receive name, usually limited to 512 bytes. If the recipient device
is not listening for messages, the datagram is lost. The Session mode establishes
a connection until broken. It guarantees delivery of messages up to 64KB
long.
WINS (Windows Internet Naming Service)
Name resolution software from Microsoft that runs under Windows NT Server
and converts NetBIOS names to IP addresses. Windows machines are assigned
NetBIOS names, which must be converted to IP addresses if the underlying
transport protocol is TCP/IP.
Windows machines identify themselves to the WINS server, so that other
Windows machines can query the server to find the IP address. Since, the
WINS server is contacted by IP address, which can be routed across subnets,
WINS allows Windows machines on one LAN segment to locate Windows machines
on other LAN segments.
When a computer is moved to another subnet and a new IP address is
assigned by DHCP, the WINS database is updated. WINS is used in a Windows
network, and both DNS and WINS are used in a mixed environment. Microsoft's
DNS server integrates the two systems. When a UNIX station wants to resolve
the name for a PC, it queries the Microsoft DNS server, which in turn queries
the WINS server if it does not already have it.
LDAP (Lightweight Directory Access Protocol)
A smaller X.500 directory service that provides authentication and certificate
management.
cdmaOne™ 2G
A brand name, trademarked and reserved for the exclusive use of CDG member
companies, that describes a complete wireless system that incorporates
the IS-95 CDMA air interface, the ANSI-41 network standard
for switch interconnection and many other standards that make up a complete
wireless system.
cdma2000 3G (or Wideband cdmaOne)
A name identifying the TIA standard for third generation technology that
is an evolutionary outgrowth of cdmaOne offering
operators who have deployed a second generation cdmaOne system a seamless
migration path that economically supports upgrade to 3G
features and services within existing spectrum allocations for both cellular
and PCS operators. The network interface defined for cdma2000 supports
the second generation network aspect of all existing operators regardless
of technology (cdmaOne, IS-136 TDMA, or GSM). This standard has been submitted
by the TIA to the ITU as part of the IMT-2000 3G process. To gracefully
migrate cdmaOne to cdma2000 capabilities offering advanced features to
the market in a flexible and timely manner, implementation has been broken
into evolutionary phases. The first phase capabilities have been defined
in a standard known as 1XRTT. Scheduled for publication in the first quarter
of 1999, 1XRTT introduces 144 kbps packet data in a mobile environment
and speeds beyond this in a fixed environment. Features available with
1XRTT are a two-fold increase in both voice capacity and standby time,
more than 300 kbps data capability, advanced packet data services, as well
as greatly extended battery life and improved sleep mode technology. All
of these capabilities will be available in an existing cdmaOne 1.25 MHz
channel. cdmaOne evolution to the complete capabilities of cdma2000 will
continue in phase two and incorporate the capabilities of 1XRTT, support
all channel sizes (5 MHz, 10 MHz, etc.), provide circuit and packet data
rates up to 2 Mbps, incorporate advanced multimedia capabilities, and include
a framework for advanced 3G voice services and vocoders, including voice
over packet and circuit data.
W-CDMA 3G (Wideband CDMA)
A name identifying the ETSI and NTT DoCoMo standards for third generation
technology submitted to the ITU as part of the IMT-2000 3G
process. This standard encompasses an air interface that uses the CDMA
technique but it is not compatible as defined for both air and network
interfaces with cdmaOne, cdma2000
or IS-136. The air interface specification is not compatible with GSM and
therefore does not support evolutionary migration.
3G (third generation communication systems)
Capable of supporting higher data rates and multimedia.
CDMA (Code Division Multiple Access)
A digital cellular phone technology from Qualcomm,
Inc., San Diego, CA, that operates in the 800MHz and 1.9GHz (PCS) bands.
CDMA phones are noted for their excellent call quality and long battery
life, and some phones can switch between both bands. CDMA uses a spread
spectrum technique that codes each digital packet and allows multiple calls
to be placed on one channel, boosting caller capacity 20 to 35 times that
of the analog network. CDMA has become widely used in North America. |