Memory Terms and Techniques
Memory: Terms and Technologies.
Contents:
|
BEDO DRAM
Short for Burst EDO DRAM.
A new type of EDO DRAM that can
process four memory addresses in one burst.
Unlike SDRAM, however, BEDO DRAM can only
stay synchronized with the CPU clock for short periods
(bursts).
Also, it can't keep up with processors whose buses run faster
than 66 MHz.
Burst Mode
A data transmission mode in which data is sent faster than
normal.
There are a number of techniques for implementing burst modes.
In a data bus, for example, a burst mode is usually implemented
by allowing a device to seize control of the bus and not
permitting other devices to interrupt.
In RAM, burst modes are implemented by
automatically fetching the next memory contents before they are
requested.
This is essentially the same technique used by disk caches.
The one characteristic that all burst modes have in common is
that they are temporary and unsustainable.
They allow faster data transfer rates than normal, but only for
a limited period of time and only under special conditions.
Main Memory
Refers to physical memory that is internal to the computer.
The word main is used to distinguish it from external mass
storage devices such as disk drives.
Another term for main memory is RAM.
The computer can manipulate only data that is in main memory.
Therefore, every program you execute and every file you
access must be copied from a storage device into main memory.
The amount of main memory on a computer is crucial because it
determines how many programs can be executed at one time and
how much data can be readily available to a program.
Because computers often have too little main memory to hold
all the data they need, computer engineers invented a
technique called swapping, in which portions of data are
copied into main memory as they are needed.
Swapping occurs when there is no room in memory for needed data.
When one portion of data is copied into memory, an
equal-sized portion is copied (swapped) out to make room.
Another technique, called virtual memory, enables a computer
to access larger amounts of data than main memory can hold at
one time, but it is a relatively slow process.
Therefore, the more memory a computer has and the more it can
avoid swapping, the faster it will be able to execute large
programs.
Now, most PCs come with a minimum of 8 megabytes of main
memory.
This is adequate for some applications, but you may need more
memory to run sophisticated applications, particularly those
that include graphics.
You can usually increase the amount of memory by inserting
extra memory in the form of chips or memory expansion boards.
Non Volatile Memory
Types of memory that retain their contents when power is
turned off.
ROM is nonvolatile, whereas
RAM is volatile.
Cache Memory
A special high-speed storage mechanism.
It can be either a reserved section of main memory or an
independent high-speed storage device.
Two types of caching are commonly used in personal computers:
- memory caching
- disk caching
A memory cache, sometimes called a cache store
or RAM cache,
is a portion of memory made of high-speed static RAM
(SRAM)
instead of the slower and cheaper dynamic
RAM (DRAM) used for
main memory.
Memory caching is effective because most programs access the
same data or instructions over and over.
By keeping as much of this information as possible in
SRAM,
the computer avoids accessing the slower
DRAM.
Some memory caches are built into the architecture of
microprocessors.
The Intel 80486 microprocessor, for example, contains an
8K memory cache, and the Pentium has a 16K cache.
Such internal caches are often called Level 1 (L1) caches.
Most modern PCs also come with external cache memory, called
Level 2 (L2) caches.
These caches sit between the CPU and the
DRAM.
Like L1 caches, L2 caches are composed of
SRAM but they are much
larger.
Disk caching works under the same principle as memory
caching, but instead of using high-speed
SRAM, a disk cache
uses conventional main memory.
The most recently accessed data from the disk (as well as
adjacent sectors) is stored in a memory buffer.
When a program needs to access data from the disk, it first
checks the disk cache to see if the data is there.
Disk caching can dramatically improve the performance
of applications, because accessing a byte of data in
RAM can
be thousands of times faster than accessing a byte on a hard
disk.
When data is found in the cache, it is called a cache hit,
and the effectiveness of a cache is judged by its hit rate.
Many cache systems use a technique known as smart caching, in
which the system can recognize certain types of frequently
used data.
The strategies for determining which information should be
kept in the cache constitute some of the more interesting
problems in computer science.
Video Memory
RAM
installed on a video adapter.
Before an image can
be sent to a display monitor it is first represented as a
bit map in an area of video memory called the frame
buffer.
The amount of video memory, therefore,
dictates the maximum resolution and color depth
available (see table below).
With a conventional video adapter, the bit map to be
displayed is first generated by the computer's
microprocessor and then sent to the frame buffer. Most
modern video adapters, however, are actually graphics
accelerators.
This means that they have they're own
microprocessor that is capable of manipulating bit
maps and graphics objects.
A small amount of memory
is reserved for these operations as well.
Because of the demands of video systems, video
memory needs to be faster than main memory. There
are many different types of video memory, including
VRAM,
WRAM,
RDRAM, and
SGRAM.
Video RAM Required for Different Resolutions.
Resolution
| 16 colors (4 bit)
| 256 colors (8 bit)
| 65 k colors (16 bit)
| 16.7 mio colors (24 bit)
|
640 x 480
| 512K
| 512K
| 1 MB
| 2 MB
|
800 x 600
| 512K
| 512K
| 1 MB
| 2 MB
|
1024 x 768
| 1 MB
| 1 MB
| 2 MB
| 4 MB
|
1152 x 1024
| 1 MB
| 2 MB
| 2 MB
| 4 MB
|
1280 x 1024
| 1 MB
| 2 MB
| 4 MB
| 4 MB
|
1600 x 1200
| 2 MB
| 2 MB
| 4 MB
| 8 MB
|
ROM
Short for read-only memory.
Computer memory on which data has been prerecorded.
Once data has been written onto a ROM chip, it cannot be
removed and can only be read.
Unlike main memory (RAM),
ROM retains its contents even when
the computer is turned off. ROM is referred to as being
nonvolatile, whereas RAM is volatile.
Most personal computers contain a small amount of ROM that
stores critical programs such as the program that boots the
computer. In addition, ROMs are used extensively in
calculators and peripheral devices such as laser printers,
whose fonts are often stored in ROMs.
A variation of a ROM is a
PROM (programmable read-only memory).
PROMs are manufactured as blank chips
on which data
can be written with a special device called a PROM programmer.
RAM
Short for random access memory.
A type of computer memory that can be accessed randomly;
that is, any byte of memory can be accessed without touching
the preceding bytes.
RAM is the most common type of memory
found in computers and other devices, such as printers.
There are two basic types of RAM:
- Dynamic RAM (DRAM)
- Static RAM (SRAM)
The two types differ in the technology they use to hold data,
dynamic RAM
being the more common type.
Dynamic RAM needs to
be refreshed thousands of times per second.
Static RAM needs
to be refreshed less often, which makes it faster; but it is
also more expensive than
dynamic RAM.
Both types of RAM are
volatile, meaning that they lose their contents when the
power is turned off.
In common usage, the term
RAM is synonymous with
main memory,
the memory available to programs. For example, a computer
with 8M RAM has approximately 8 million
bytes of memory that programs can use.
In contrast, ROM (read-only memory) refers
to special memory used to store programs that boot the
computer and perform diagnostics.
Most personal computers
have a small amount of
ROM (a few thousand bytes).
In fact,
both types of memory
(ROM
and RAM) allow random access.
To be
precise, therefore, RAM should be referred
to as read/write
RAM and
ROM as read-only RAM.
RAMDAC
Short for Random Access Memory Digital-to-Analog Converter.
A single chip on video adapter cards.
The RAMDAC's role is to convert digitally encoded images into
analog signals that can be displayed by a monitor.
A RAMDAC actually consists of four different components
- SRAM to store the color map
- digital-to-analog converters (DAC) for monitor's red electron gun
- digital-to-analog converters (DAC) for monitor's green electron gun
- digital-to-analog converters (DAC) for monitor's blue electron gun
PROM
Short for programmable read-only memory.
A
PROM
is a memory chip on which data can be written only once.
Once a program has been written onto a
PROM, it remains there
forever.
Unlike main memory,
PROMs retain their contents when the
computer is turned off.
The difference between a
PROM and a
ROM (read-only memory) is
that a
PROM is manufactured as blank memory,
whereas a ROM is
programmed during the manufacturing process.
To write data
onto a
PROM
chip, you need a special device called a PROM
programmer or PROM burner.
The process of programming a
PROM
is sometimes called burning the PROM.
An EPROM (erasable programmable read-only memory) is a
special type of
PROM
that can be erased by exposing it to
ultraviolet light.
Once it is erased, it can be reprogrammed.
An
EEPROM
is similar to a
PROM, but requires only electricity
to be erased.
EEPROM
Short for electrically erasable programmable read-only
memory.
An
EEPROM
is a special type of
PROM
that can be erased by exposing it to an electrical charge.
Like other types of
PROM,
EEPROM
retains its contents even when the
power is turned off.
Also like other types of
ROM,
EEPROM
is not as fast as
RAM.
EEPROM
is similar to flash memory (sometimes called flash EEPROM).
The principal difference is that
EEPROM requires
data to be written or erased one byte at a time whereas flash
memory allows data to be written or erased in blocks.
This
makes flash memory faster.
NVRAM
Short for Non-Volatile Random Access Memory.
Any type of memory that retains its contents when power is
turned off.
Also a type of
SRAM
that is made non-volatile by connecting it
to a constant power source such as a battery.
Another type of NVRAM uses
EEPROM
chips to save its contents
when power is turned off.
In this case, NVRAM is composed of a combination of
SRAM and
EEPROM chips.
MDRAM
Short for Multibank DRAM.
A relatively new memory technology
developed by MoSys Inc.
MDRAM utilizes small banks of
DRAM
(32 KB each) in an array, where each bank has its own I/O
port that feeds into a common internal bus.
Because of this
design, data can be read or written to multiple banks
simultaneously, which makes it much faster than conventional
DRAM.
Another advantage of MDRAM is that memory can be configured
in smaller increments, which can reduce the cost of some
components.
For example, it's possible to produce MDRAM chips
with 2.5 MB, which is what is required by video adapters for
24-bit color at a resolution of 1,024x768.
With conventional
memory architectures, it's necessary to jump all the way to 4
MB. Currently, MDRAM is used in some video adapters and
graphics accelerators.
FPM RAM
Short for Fast Page Mode RAM.
A type of Dynamic RAM (
DRAM) that allows faster access to
data in the same row or page.
Page-mode memory works by eliminating the need for a row address
if data is located in the row previously accessed.
It is sometimes called page mode memory.
FPM RAM is being replaced by newer types of memory, such as
Extended Data Out (EDO) DRAM,
BEDO DRAM, and
SDRAM.
EDO DRAM
Short for Extended Data Output Dynamic Random Access Memory.
A type of DRAM that is faster than
conventional DRAM.
Unlike conventional DRAM which only allows
one byte to be read at a time.
EDO DRAM can copy an entire block of memory to its internal
cache.
While the processor is accessing this cache,
the memory can collect a new block to send.
Note that EDO DRAM is faster than conventional
DRAM only if the cache controller
supports a transfer mode known as
pipeline burst.
Nearly all PCs that support EDO DRAM are equipped with such a
controller, but some Macintosh models are not.
DRAM
Short for dynamic random access memory.
A type of memory used in most personal computers.
See also SRAM.
DDR SDRAM
Short for Double Data Rate-Synchronous DRAM.
A type of SDRAM that supports data
transfers on both edges of each clock cycle, effectively
doubling the memory chip's data throughput.
DDR-SDRAM is also called SDRAM II.
RDRAM
Short for Rambus DRAM.
A type of memory (DRAM) developed by
Rambus, Inc.
Whereas
the fastest current memory technologies used by PCs
(SDRAM)
can deliver data at a maximum speed of about 100 MHz,
RDRAM transfers data at up to 600 MHz.
In 1997, Intel announced that it would license the Rambus
technology for use on its future motherboards, thus making it
the likely de facto standard for memory architectures.
However, a consortium of computer vendors is working on an
alternative memory architecture called SyncLink DRAM
(SLDRAM).
RDRAM is already being used in place of
VRAM in some graphics
accelerator boards, but it is not expected to be used for the
main memory of PCs until 1998 or 1999.
Intel and Rambus are
also working a new version of RDRAM, called nDRAM, that will
support data transfer speeds at up to 1,600 MHz.
SDRAM
Short for Synchronous DRAM.
A new type of
DRAM that can run at much higher clock speeds than
conventional memory.
SDRAM actually synchronizes itself with the CPU's bus and is
capable of running at 100 MHz, about three times faster than
conventional
FPM RAM, and about twice as fast
EDO DRAM and
BEDO DRAM.
SDRAM is replacing
EDO DRAM in many newer computers.
Today's fastest Pentium systems use CPU buses running at 100
MHz, so SDRAM can keep up with them, though barely.
Future
PCs, however, are expected to have CPU buses running at 200
MHz or faster.
SDRAM is not expected to support these high
speeds which is why new memory technologies, such as
RDRAM and
SLDRAM, are being developed.
SGRAM
Short for Synchronous Graphic Random Access Memory.
A type of
DRAM
used increasingly on video adapters and graphics accelerators.
Like
SDRAM,
SGRAM can synchronize itself with
the CPU bus clock up to speeds of 100 MHz.
In addition, SGRAM uses several other techniques, such as
masked writes and block writes, to increase bandwidth for
graphics-intensive functions.
Unlike
VRAM and
WRAM, SGRAM is single-ported.
However, it can
open two memory pages at once, which simulates the dual-port
nature of other video RAM technologies.
SLDRAM
Short for SyncLink DRAM.
A new type of memory being developed by a consortium of computer
manufacturers called the Synclink Consortium.
SLDRAM is competing with Rambus memory
(RDRAM) as the future PC
memory architecture.
But whereas actual
RDRAM chips are already
in use on video boards and other devices, SLDRAM is still
on paper only.
Moreover, Intel is backing
RDRAM, which reduces the odds of SLDRAM
becoming an important technology.
SRAM
Short for static random access memory.
SRAM is a type of memory that is faster and more reliable than
the more common
DRAM (dynamic RAM).
The term static is derived from the fact that it needs to be
refreshed less often than
dynamic RAM.
While DRAM supports access times of
about 60 nanoseconds,
SRAM can give access times as low as 10 nanoseconds.
In addition, its cycle time is much shorter than that of
DRAM
because it does not need to pause between accesses.
Unfortunately, it is also much more expensive to produce than
DRAM.
Due to its high speed, SRAM is often used only as a memory cache.
VRAM
Short for video RAM.
VRAM is special-purpose memory used by video adapters.
Unlike conventional
RAM, VRAM can be accessed by two different
devices simultaneously.
This enables a monitor to access the VRAM for screen updates at
the same time that a graphics processor provides new data.
VRAM yields better graphics performance but is more expensive
than normal RAM.
WRAM
Short for Window RAM.
A type of
RAM
developed by Samsung Electronics that supports
two ports.
This enables a video adapter to fetch the contents
of memory for display at the same time that new bytes are
being pumped into memory.
This results in much faster display than is possible with
conventional single-port
RAM.
WRAM is similar to
VRAM, but achieves even faster performance
at less cost.
Pipeline Burst Cache
A type of memory cache built into many modern
DRAM designs.
Pipeline burst caches use two techniques:
- burst mode that pre-fetches memory contents before they
are requested
- pipelining so that one memory value can be accessed in the
cache at the same time that another memory value is accessed
in DRAM
The purpose of pipeline burst caches is to minimize
wait states so that memory can be accessed as fast a possible
by the microprocessor.
An alternative DRAM
design that provides even faster DRAM
access is Synchronous DRAM (SDRAM).
SDRAM also uses a burst mode, but because
memory accesses are synchronized with the microprocessor bus,
pipelining is not necessary.
Pipelining
(1)
A technique used in advanced microprocessors where the
microprocessor begins executing a second instruction before the
first has been completed.
That is, several instructions are in the pipeline simultaneously,
each at a different processing stage.
The pipeline is divided into segments and each segment can
execute its operation concurrently with the other segments.
When a segment completes an operation, it passes the result
to the next segment in the pipeline and fetches the next
operation from the preceding segment.
The final results of each instruction emerge at the end of the
pipeline in rapid succession.
Although formerly a feature only of high-performance and RISC
-based microprocessors, pipelining is now common in
microprocessors used in personal computers.
Intel's Pentium chip, for example, uses pipelining to execute as
many as four instructions simultaneously.
Pipelining is called pipeline processing.
(2)
A similar technique used in
DRAM, in which the memory loads the
requested memory contents into a small cache composed of
SRAM
and then immediately begins fetching the next memory contents.
This creates a two-stage pipeline, where data is read from or
written to
SRAM in one stage, and data is read from or
written to memory in the other stage.
DRAM
pipelining is usually combined with another performance
technique called burst mode.
The two techniques together are called a
pipeline burst cache.
Thanks for joining.
This is the end of the show.
Juergen Theis, IBM Germany
e-mail: jtheis@vnet.ibm.com
Last update: 25.03.98