Micro Finall

download Micro Finall

of 14

Transcript of Micro Finall

  • 8/8/2019 Micro Finall

    1/14

    Computer Memory

    Computer memory is quite like Human memory. Demonstration for how the

    brain is like a computer.

    Information Processing

    Parallel Processing

    Some scientists use computers to simulate how a brain might work.Instead of focusing on the chemical aspects of memory, they simulate the neuralconnections made in the processing of information.

    Computer memory can be any device, such as a silicon chip or a hard disk, usedto store information. In computers of the past, vacuum tubes, cores and drumshave been used. Today, computer chips are used.

    Of course, computers' memory capabilities are increasing as technology advances.

    Computer chips' capabilities have expanded. Other computer memory devices include

    CD-ROMs, floppy disks, tapes, zip drives, DVDs, etc.

    Memory

    The term "memory" applies to any electronic component capable oftemporarily storing data. There are two main categories of memories.

    Internal Memory:

    Internal memory that temporarily memorises data while programs arerunning. Internal memory uses microconductors, i.e. fast specialised electroniccircuits. Internal memory corresponds to what we call random access memory(RAM).

    http://library.thinkquest.org/C0110291/basic/computer/process.phphttp://library.thinkquest.org/C0110291/basic/computer/parallel.phphttp://library.thinkquest.org/C0110291/basic/computer/pastcomputers.phphttp://en.kioskea.net/contents/pc/#RAMhttp://library.thinkquest.org/C0110291/basic/computer/process.phphttp://library.thinkquest.org/C0110291/basic/computer/parallel.phphttp://library.thinkquest.org/C0110291/basic/computer/pastcomputers.phphttp://en.kioskea.net/contents/pc/#RAM
  • 8/8/2019 Micro Finall

    2/14

  • 8/8/2019 Micro Finall

    3/14

    Memory Devices

    Computers and many electronic gadgets usually rely on stored informationwhich is mainly data which can be used to direct circuit actions. The digital

    information is stored in memory devices. Memories can be divided into 2categories based on what memory cells can be accessed at a given instant. SAM(Sequentially Access Memory) is accessed by stepping through each memorylocation until the desired location is reached. Magnetic tape is an example ofSAM.

    The second category of memory devices is called RAM (Random AccessMemory) where the memory can be randomly accessed at any instant, withouthaving to step through each memory location. It is generally faster to access aRAM compared to SAM. Most of the electronics gadgets memory is of RAM type.

    Random Access Memory :Random access memory, generally called RAM is the system's main

    memory, i.e. it is a space that allows you to temporarily store data when aprogram is running.

    Unlike data storage on an auxiliary memory such as a hard drive, RAM isvolatile, meaning that it only stores data as long as it supplied with electricity.Thus, each time the computer is turned off, all the data in the memory areirremediably erased.

    Types of RAM :

    The RAM family includes two important memory devices: static RAM(SRAM) and dynamic RAM (DRAM). The primary difference between them is thelifetime of the data they store. SRAM retains its contents as long as electricalpower is applied to the chip. If the power is turned off or lost temporarily, itscontents will be lost forever. DRAM, on the other hand, has an extremely short

    http://eml2%28%27e/e8/','DDR2_ram_mounted.jpg')
  • 8/8/2019 Micro Finall

    4/14

    data lifetime-typically about four milliseconds. This is true even when power isapplied constantly.

    In short, SRAM has all the properties of the memory you think of when youhear the word RAM. Compared to that, DRAM seems kind of useless. By itself, itis. However, a simple piece of hardware called a DRAM controller can be used to

    make DRAM behave more like SRAM. The job of the DRAM controller is toperiodically refresh the data stored in the DRAM. By refreshing the data before itexpires, the contents of memory can be kept alive for as long as they areneeded. So DRAM is as useful as SRAM after all.

    NVRAM chip consists of both RAM and ROM. During power on reset, thecontents of the ROM are copied to RAM. Before the power turns off, the systemwill copy the entire contents of the RAM into ROM for non volatile storage. TheRAM in an NVRAM is called shadow RAM. NVRAM fills the gap between easilywritten memory and non volatile memory.

    Read-Only Memory :

    Read-only memory, called ROM, is a type of memory that allows you tokeep the information contained on it even when the memory is no longerreceiving electricity. Basically, this type of memory only has read-only access.However, it is possible to save information in some types of ROM memory.

    ROM is non volatile in that its contents are not lost when power to it isremoved. All ROMs can be programmed at least once. Mask ROMs areprogrammed by having "1"s and "0"s etched into their semiconductors at the timeof manufacturing. Programmable ROM (PROM) can be written aftermanufacturing by electrically burning specific transistors or diodes in the memoryarray. EPROM can be erased and reprogrammed by using ultraviolet light.

    EEPROM (electronically eraseable PROM) data can be erasedelectronically but it takes longer time compared to RAM. Read and write time forRAM is almost the same but PROM has slower write times. PROM must beerased before they can be reprogrammed and it often needs a higherprogramming voltage than its operating voltage.

    ROM is usually used to store data or programs that do not changefrequently and must still be there when power supply cuts off.

    http://eml2%28%27e/eb/','Seagate_Hard_Disk.jpg')
  • 8/8/2019 Micro Finall

    5/14

    Types of ROM :

    Memories in the ROM family are distinguished by the methods used towrite new data to them (usually called programming), and the number of timesthey can be rewritten. This classification reflects the evolution of ROM devices

    from hardwired to programmable to erasable-and-programmable. A commonfeature of all these devices is their ability to retain data and programs forever,even during a power failure.

    The very first ROMs were hardwired devices that contained apreprogrammed set of data or instructions. The contents of the ROM had to bespecified before chip production, so the actual data could be used to arrange thetransistors inside the chip. Hardwired memories are still used, though they arenow called "masked ROMs" to distinguish them from other types of ROM. Theprimary advantage of a masked ROM is its low production cost. Unfortunately,the cost is low only when large quantities of the same ROM are required.

    One step up from the masked ROM is the PROM (programmable ROM),

    which is purchased in an unprogrammed state. If you were to look at the contentsof an unprogrammed PROM, you would see that the data is made up entirely of1's. The process of writing your data to the PROM involves a special piece ofequipment called a device programmer. The device programmer writes data tothe device one word at a time by applying an electrical charge to the input pins ofthe chip. Once a PROM has been programmed in this way, its contents cannever be changed. If the code or data stored in the PROM must be changed, thecurrent device must be discarded. As a result, PROMs are also known as one-time programmable (OTP) devices.

    An EPROM (erasable-and-programmable ROM) is programmed inexactly the same manner as a PROM. However, EPROMs can be erased and

    reprogrammed repeatedly. To erase an EPROM, you simply expose the deviceto a strong source of ultraviolet light. (A window in the top of the device allowsthe light to reach the silicon.) By doing this, you essentially reset the entire chipto its initial--unprogrammed--state. Though more expensive than PROMs, theirability to be reprogrammed makes EPROMs an essential part of the softwaredevelopment and testing process.

    Hybrids :

    As memory technology has matured in recent years, the line betweenRAM and ROM has blurred. Now, several types of memory combine features ofboth. These devices do not belong to either group and can be collectivelyreferred to as hybrid memory devices. Hybrid memories can be read and writtenas desired, like RAM, but maintain their contents without electrical power, justlike ROM. Two of the hybrid devices, EEPROM and flash, are descendants ofROM devices. These are typically used to store code. The third hybrid, NVRAM,is a modified version of SRAM. NVRAM usually holds persistent data.

  • 8/8/2019 Micro Finall

    6/14

    EEPROMs are electrically-erasable-and-programmable. Internally, theyare similar to EPROMs, but the erase operation is accomplished electrically,rather than by exposure to ultraviolet light. Any byte within an EEPROM may beerased and rewritten. Once written, the new data will remain in the deviceforever--or at least until it is electrically erased. The primary tradeoff for this

    improved functionality is higher cost, though write cycles are also significantlylonger than writes to a RAM. So you wouldn't want to use an EEPROM for yourmain system memory.

    Flash memory combines the best features of the memory devicesdescribed thus far. Flash memory devices are high density, low cost, nonvolatile,fast (to read, but not to write), and electrically reprogrammable. Theseadvantages are overwhelming and, as a direct result, the use of flash memoryhas increased dramatically in embedded systems. From a software viewpoint,flash and EEPROM technologies are very similar. The major difference is thatflash devices can only be erased one sector at a time, not byte-by-byte. Typicalsector sizes are in the range 256 bytes to 16KB. Despite this disadvantage, flash

    is much more popular than EEPROM and is rapidly displacing many of the ROMdevices as well.The third member of the hybrid memory class is NVRAM (non-volatile

    RAM). Nonvolatility is also a characteristic of the ROM and hybrid memoriesdiscussed previously. However, an NVRAM is physically very different from thosedevices. An NVRAM is usually just an SRAM with a battery backup. When thepower is turned on, the NVRAM operates just like any other SRAM. When thepower is turned off, the NVRAM draws just enough power from the battery toretain its data. NVRAM is fairly common in embedded systems. However, it isexpensive--even more expensive than SRAM, because of the battery--so itsapplications are typically limited to the storage of a few hundred bytes of system-

    critical information that can't be stored in any better way.

  • 8/8/2019 Micro Finall

    7/14

    Flash Memory :

    Flash memory is a compromise between RAM-type memories and ROMmemories. Flash memory possesses the non-volatility of ROM memories whileproviding both read and write access However, the access times of flash

    memories are longer than the access times of RAM.

    Flash memory

    A USB flash drive. The chip on the left is the flash memory. Themicrocontrolleris on the right.

    Flash memory is a non-volatilecomputer memory that can be electrically

    erased and reprogrammed. It is a technology that is primarily used in memorycards and USB flash drives for general storage and transfer of data betweencomputers and other digital products. It is a specific type of EEPROM(Electrically Erasable Programmable Read-Only Memory) that is erased andprogrammed in large blocks; in early flash the entire chip had to be erased atonce. Flash memory costs far less than byte-programmable EEPROM andtherefore has become the dominant technology wherever a significant amount ofnon-volatile, solid state storage is needed. Example applications include PDAs(personal digital assistants), laptop computers, digital audio players, digitalcameras and mobile phones. It has also gained popularity in the game consolemarket, where it is often used instead of EEPROMs or battery-powered SRAM

    for game save data.Flash memory is non-volatile, which means that no power is needed tomaintain the information stored in the chip. In addition, flash memory offers fastread access times (although not as fast as volatile DRAM memory used for mainmemory in PCs) and better kinetic shock resistance than hard disks. Thesecharacteristics explain the popularity of flash memory in portable devices.Another feature of flash memory is that when packaged in a "memory card," it is

    http://en.wikipedia.org/wiki/USB_flash_drivehttp://en.wikipedia.org/wiki/Microcontrollerhttp://en.wikipedia.org/wiki/Non-volatile_memoryhttp://en.wikipedia.org/wiki/Non-volatile_memoryhttp://en.wikipedia.org/wiki/Computer_storagehttp://en.wikipedia.org/wiki/Memory_cardhttp://en.wikipedia.org/wiki/Memory_cardhttp://en.wikipedia.org/wiki/USB_flash_drivehttp://en.wikipedia.org/wiki/EEPROMhttp://en.wikipedia.org/wiki/Solid_state_(electronics)http://en.wikipedia.org/wiki/Personal_digital_assistanthttp://en.wikipedia.org/wiki/Digital_audio_playerhttp://en.wikipedia.org/wiki/Digital_camerahttp://en.wikipedia.org/wiki/Digital_camerahttp://en.wikipedia.org/wiki/Digital_camerahttp://en.wikipedia.org/wiki/Mobile_phonehttp://en.wikipedia.org/wiki/EEPROMhttp://en.wikipedia.org/wiki/Static_RAMhttp://en.wikipedia.org/wiki/Non-volatile_memoryhttp://en.wikipedia.org/wiki/Access_timehttp://en.wikipedia.org/wiki/Dynamic_random_access_memoryhttp://en.wikipedia.org/wiki/Hard_diskhttp://en.wikipedia.org/wiki/File:DSCN0411.JPGhttp://en.wikipedia.org/wiki/USB_flash_drivehttp://en.wikipedia.org/wiki/Microcontrollerhttp://en.wikipedia.org/wiki/Non-volatile_memoryhttp://en.wikipedia.org/wiki/Computer_storagehttp://en.wikipedia.org/wiki/Memory_cardhttp://en.wikipedia.org/wiki/Memory_cardhttp://en.wikipedia.org/wiki/USB_flash_drivehttp://en.wikipedia.org/wiki/EEPROMhttp://en.wikipedia.org/wiki/Solid_state_(electronics)http://en.wikipedia.org/wiki/Personal_digital_assistanthttp://en.wikipedia.org/wiki/Digital_audio_playerhttp://en.wikipedia.org/wiki/Digital_camerahttp://en.wikipedia.org/wiki/Digital_camerahttp://en.wikipedia.org/wiki/Mobile_phonehttp://en.wikipedia.org/wiki/EEPROMhttp://en.wikipedia.org/wiki/Static_RAMhttp://en.wikipedia.org/wiki/Non-volatile_memoryhttp://en.wikipedia.org/wiki/Access_timehttp://en.wikipedia.org/wiki/Dynamic_random_access_memoryhttp://en.wikipedia.org/wiki/Hard_disk
  • 8/8/2019 Micro Finall

    8/14

    enormously durable, being able to withstand intense pressure, extremes oftemperature, and even immersion in water.

    Although technically a type ofEEPROM, the term "EEPROM" is generallyused to refer specifically to non-flash EEPROM which is erasable in small blocks,typically bytes. Because erase cycles are slow, the large block sizes used in

    flash memory erasing give it a significant speed advantage over old-styleEEPROM when writing large amounts of data.

    Memory card :

    A memory card or flash memory card is a solid-state electronic flashmemorydata storage device capable of storing digital contents. These are mainlyused with digital cameras, handheld and Mobile computers, mobile phones,music players, video game consoles, and otherelectronics. They offer high re-record-ability, power-free storage, small form factor, and rugged environmentalspecifications. There are also non-solid-state memory cards that do not use flash

    memory, and there are different types of flash memory

    There are many different types of memory cards and jobs they are used for.Some common places include in digital cameras, game consoles, cell phones,and industrial applications. PC card (PCMCIA) were among first commercialmemory card formats (type I cards) to come out in the 1990s, but are now onlymainly used in industrial applications and for I/O jobs (using types I/II/III), as aconnection standard for devices (such as a modem). Also in 1990s, a number ofmemory card formats smaller than PC Card came out, including CompactFlash,SmartMedia, and Miniature Card. In other areas, tiny embedded memory cards(SID) were used in cell phones, game consoles started using proprietary memorycard formats, and devices like PDAs and digital music players started usingremovable memory cards.

    Nowadays, most new PCs have built-in slots for a variety of memorycards; Memory Stick, CompactFlash, SD, etc. Some digital gadgets support morethan one memory card to ensure compatibility.

    http://en.wikipedia.org/wiki/EEPROMhttp://en.wikipedia.org/wiki/Flash_memoryhttp://en.wikipedia.org/wiki/Flash_memoryhttp://en.wikipedia.org/wiki/Data_storage_devicehttp://en.wikipedia.org/wiki/Digital_camerahttp://en.wikipedia.org/wiki/Personal_Digital_Assistanthttp://en.wikipedia.org/wiki/Mobilehttp://en.wikipedia.org/wiki/Musichttp://en.wikipedia.org/wiki/Video_game_consolehttp://en.wikipedia.org/wiki/Electronicshttp://en.wikipedia.org/wiki/Electronicshttp://en.wikipedia.org/wiki/Form_factorhttp://en.wikipedia.org/wiki/PC_cardhttp://en.wikipedia.org/wiki/SmartMediahttp://en.wikipedia.org/wiki/Personal_digital_assistanthttp://en.wikipedia.org/wiki/File:Memory_cards_shrinking_down.jpghttp://en.wikipedia.org/wiki/EEPROMhttp://en.wikipedia.org/wiki/Flash_memoryhttp://en.wikipedia.org/wiki/Flash_memoryhttp://en.wikipedia.org/wiki/Data_storage_devicehttp://en.wikipedia.org/wiki/Digital_camerahttp://en.wikipedia.org/wiki/Personal_Digital_Assistanthttp://en.wikipedia.org/wiki/Mobilehttp://en.wikipedia.org/wiki/Musichttp://en.wikipedia.org/wiki/Video_game_consolehttp://en.wikipedia.org/wiki/Electronicshttp://en.wikipedia.org/wiki/Form_factorhttp://en.wikipedia.org/wiki/PC_cardhttp://en.wikipedia.org/wiki/SmartMediahttp://en.wikipedia.org/wiki/Personal_digital_assistant
  • 8/8/2019 Micro Finall

    9/14

    Other Memory Devices :

    There are several other kinds of memory devices available for use withboth general-purpose and embedded computer systems. They include floppydisks, hard drives, CD-ROM disks and many others. When using these types of

    devices however, an additional embedded controller generally handles thecommunication between the main processor and the hardware device. Of course,this is a kind of embedded programming, but is an area that we will only brieflydiscuss, since many embedded controllers already have this kind of support built-in, or included on the main board of the system.

    Memory I nterfaces

    Interface :

    A boundary across which two independent systems meet and act on orcommunicate with each other. In computer technology, there are several types ofinterfaces.

    User interface - the keyboard, mouse, menus of a computer system.The user interface allows the user to communicate with the operating system.Also see GUI.

    Software interface - the languages and codes that the applications use

    to communicate with each other and with the hardware. Hardware interface - the wires, plugs and sockets that hardwaredevices use to communicate with each other.

    Memory Interfaces :

    Lattice provides a wide range of high-performance interface solutions forthe latest memory technologies. These solutions combine innovative silicon withIntellectual Property (IP) cores to provide robust solutions for networking

    applications.

    A memory interface is disclosed for accessing a plurality in memoryregions. The interface includes a register which stores a number of memoryrequest signals received from a processor or the like. The memory interfaceincludes circuitry for detecting which memory region each memory request refersto and also which page within that memory region is required to be accessed.Using the information contained in the register, the memory interface is able to

    http://www.webopedia.com/TERM/I/GUI.htmlhttp://www.webopedia.com/TERM/I/GUI.html
  • 8/8/2019 Micro Finall

    10/14

    determine which page within a memory region will be required to be accessedafter the currently open page is closed. The memory interface can detect thisinformation a number of memory requests in advance. Thus the memoryinterface is able to provide the necessary control instructions to initiate theopening of the subsequently required page within a memory region so that when

    the memory request requiring access to this page is serviced, there is no delay inopening the page. The memory interface is arranged so that a page within a firstmemory region can be opened while a page within a second memory is beingactually accessed.

    Features :

    LatticeSC FPGA devices provide full-featured embedded high-speedmemory controllers supporting DDRI/II SDRAM, QDR I/II SRAM, andRLDRAM I/II memory devices.

    LatticeECP3/ECP2/M/ECP/XP2/XP FPGA devices provide dedicatedresources to align DQ and DQS signals, multiplex/de-multiplex to and fromdouble data rate, and transfer data from the DQS clock domain to the systemclock domain.

    LatticeECP3 devices provide support for DDR3 Read and Writeleveling to adjust for PCB route delay on read and write data transfers.

    Lattice ORSPI4 FPSC contains an embedded QDR II memoryinterface providing 20+ Gbps bandwidth w/simple FIFO interface to FPGA.

    Through the ispLeverCORE program, Lattice offers a variety of IPcores and Reference Designs for popular memory interfaces.

    External Memory Interface :

    An External Memory Interface is a bus protocol for communication froman integrated circuit, such as a microprocessor, to an external memory devicelocated on a circuit board. The memory is referred to as external because it is notcontained within the internal circuitry of the integrated circuit and thus isexternally located on the circuit board.

    Some common External Memory Interfaces include:

    DDR DDR2 GDDR

    http://www.latticesemi.com/products/intellectualproperty/index.cfm?jsessionid=f03039beefb9d71d5edc1c3c5536355e3f74http://www.latticesemi.com/products/intellectualproperty/index.cfm?jsessionid=f03039beefb9d71d5edc1c3c5536355e3f74http://en.wikipedia.org/wiki/Integrated_circuithttp://en.wikipedia.org/wiki/Microprocessorhttp://en.wikipedia.org/w/index.php?title=External_memory_device&action=edit&redlink=1http://en.wikipedia.org/wiki/Circuit_boardhttp://en.wikipedia.org/wiki/DDRhttp://en.wikipedia.org/wiki/DDR2http://en.wikipedia.org/wiki/GDDRhttp://www.latticesemi.com/products/intellectualproperty/index.cfm?jsessionid=f03039beefb9d71d5edc1c3c5536355e3f74http://www.latticesemi.com/products/intellectualproperty/index.cfm?jsessionid=f03039beefb9d71d5edc1c3c5536355e3f74http://en.wikipedia.org/wiki/Integrated_circuithttp://en.wikipedia.org/wiki/Microprocessorhttp://en.wikipedia.org/w/index.php?title=External_memory_device&action=edit&redlink=1http://en.wikipedia.org/wiki/Circuit_boardhttp://en.wikipedia.org/wiki/DDRhttp://en.wikipedia.org/wiki/DDR2http://en.wikipedia.org/wiki/GDDR
  • 8/8/2019 Micro Finall

    11/14

    Common Flash memory Interface :

    The Common Flash memory Interface (CFI) is an open standard jointlydeveloped by AMD, Intel, Sharp and Fujitsu. An overview about the specificationis available at AMD. It is an open standard, which means it is freely

    implementable by all flash memory vendors, and has been approved by the non-volatile memory subcommittee of JEDEC[1]. The idea behind was theinterchangeability of current and future flash memory devices offered by differentvendors. The developer is able to use one driver for different flash products byreading identifying information out of the flash chip itself as can be read hereIntel.

    That information contains:

    Memory size

    Byte and word configuration

    Block configuration Voltages and timings

    Benefits of this concept are:

    Basically no or little information about flash device has to be stored intables within system software

    Possible to use lower cost flash memory devices as they becomeavailable, without rewriting code

    Adapting current software systems shall be done more easily andquickly than before

    Memory interface device andmethod f or Accessing Memories

    A memory interface is disclosed for accessing a plurality in memoryregions. The interface includes a register which stores a number of memoryrequest signals received from a processor or the like. The memory interface

    includes circuitry for detecting which memory region each memory request refersto and also which page within that memory region is required to be accessed.Using the information contained in the register, the memory interface is able todetermine which page within a memory region will be required to

    http://en.wikipedia.org/wiki/AMDhttp://en.wikipedia.org/wiki/Intelhttp://en.wikipedia.org/wiki/Sharp_Corporationhttp://en.wikipedia.org/wiki/Fujitsuhttp://www.amd.com/us-en/assets/content_type/DownloadableAssets/cfi_r20.pdfhttp://en.wikipedia.org/wiki/Flash_memoryhttp://en.wikipedia.org/wiki/JEDEChttp://en.wikipedia.org/wiki/JEDEChttp://en.wikipedia.org/wiki/Common_Flash_Memory_Interface#cite_note-0%23cite_note-0http://en.wikipedia.org/wiki/AMDhttp://en.wikipedia.org/wiki/Intelhttp://en.wikipedia.org/wiki/Sharp_Corporationhttp://en.wikipedia.org/wiki/Fujitsuhttp://www.amd.com/us-en/assets/content_type/DownloadableAssets/cfi_r20.pdfhttp://en.wikipedia.org/wiki/Flash_memoryhttp://en.wikipedia.org/wiki/JEDEChttp://en.wikipedia.org/wiki/Common_Flash_Memory_Interface#cite_note-0%23cite_note-0
  • 8/8/2019 Micro Finall

    12/14

    Choosing and using microprocessor memory interfaces

    Microprocessor-based systems are ideal for executing an essentially

    infinite number of tasks. The host microprocessors support a limited set of

    instructions that can combine to produce incredibly complex software programs.

    In other words, a microprocessor Is a piece of hardware designed to be asgeneral-purpose as is feasible, to target as many applications as possible.

    Moore's Law states that microprocessors will double in complexity roughly

    every two years. This remarkably accurate prediction has resulted in the latest

    multi core processors that enable the convergence of cutting-edge technologies.However, a microprocessor does not make up a complete system. In

    reality, the supporting components in a system are just as important as themicroprocessor itself when determining overall capabilities and performance.Just as microprocessors evolve with time into faster and more efficient devices,supporting components are also evolving to include more complex functions andhigher performance interfaces.

    PCIE (Peripheral Component Interconnect Express) is quickly becomingthe peripheral interconnect of choice, because the slower PCI and AGP(Accelerated Graphics Port) buses place bottlenecks on system performance.For similar reasons, DDR2 (Double Data Rate 2) is slowly taking hold as ageneral-purpose memory to overcome its slower predecessor, DDR. A system'smemory interface can affect performance more than any other system-levelinterface, and no interface offers more choices and configurations.

    At the system level, PCIe interfaces offer configurable options in the form

    of data rates and lane widths (one, two, four, and, in some cases, eight lanes). Incontrast, DDR2 interfaces frequently have widths of 4 to 256 bits and offer amultitude of capacity, data-rate, and core-timing-performance permutations. Addthe variety of available memory technologies, and system designers end up withthe daunting task of finding an optimal configuration for their systems.

    DDR, DDR2, RDRAM (Rambus dynamic-random-access memory),GDDR1/2/3 (graphics DDR1/2/3), and XDR (extreme-data-rate) DRAM are allexamples of memory technologies that designers frequently use, and each onehas its own set of advantages and drawbacks. To add to this already populoustechnology space, some memory companies have decided to produce specialty

    memories that offer superior performance for specially targeted markets.RLDRAM (reduced-latency DRAM) and FCRAM (fast-cycle RAM), for example,are two technologies specifically optimized for network-processor manufacturersthat require fast internal DRAM cycle times. XDR2 is a new memory technologyfrom Rambus that incorporates micro-threading and offers high efficiency forgraphics, networking, and consumer-electronics applications.

  • 8/8/2019 Micro Finall

    13/14

    Depending on system needs, designers must choose a memorytechnology and configuration that minimizes the overall system cost andmaximizes performance. Typically, designers optimize a memory system for anycombination of cost versus capacity, peak bandwidth, efficiency, and system-level restrictions. Although capacity seems straightforward enough, when adding

    more devices to get more memory, designers must carefully consider how to addthe memory devices into the system. For example, adding capacity involves asystem-level trade-off, because the added memory devices draw more power.

    Enterprise servers and supercomputers are often optimized, at least inpart, to contain high-capacity-memory systems. The large and complex programsand data sets that servers and supercomputers often access benefit from morecapacity. Users access instructions and data stored in rapid-access areas, suchas DRAM subsystems, more quickly than those stored in low-speed, high-latencystorage areas, such as hard-disk drives. Therefore, in addition to capacity, thesesystems are also clearly sensitive to peak bandwidth.

    Two ways exist to increase the peak bandwidth of a memory system:increased bus width and increased data rate. The latter involves increasing therate at which data transfers on each data link, and the former involves increasingthe number of data links in the memory system to obtain a higher total aggregatebandwidth. For example, to obtain a total aggregate bandwidth of 12.8 Gbytes/sec, a designer could opt for a 128-bit-wide DDR2 system running at an800-MHz data rate or a 16-bit-wide XDR system running at a 6.4-GHz data rate.

    Most of today's memory technologies provide the capability of achievinghigh total aggregate data rates, but different applications stress the memory

    system in different ways. The architectural features of any given memorytechnology dictate its efficiency for a particular application and its correspondingmemory-access requirements. The efficiency of a memory subsystem is definedas the percentage of a system's total aggregate bandwidth that provides usefuldata to and from the host microprocessor and is the reason memory vendorshave added specialty memories to their product portfolios.

  • 8/8/2019 Micro Finall

    14/14