Systems Architecture, Fifth Edition

1
SCSC 311 Information Systems hardware and
software
2
(No Transcript)
3
Chapter Objectives

• Characteristics of primary and secondary storage
• Implementation of primary storage
• Memory allocation schemes
• Alternative secondary storage technology
• Storage device performance
• Choose appropriate secondary storage technologies
  and devices

4
Storage Devices

• Storage devices consist of
• Storage medium device or substance holds data
• Read/write mechanism read/write data to/from the
  storage medium
• e.g. RAM, HD, magnetic tape, CD, USB flash memory
  
• Device controller provides the interface between
  the storage devices and system bus
• Two main types
• Primary storage devices
• Support immediate execution of programs
• Secondary storage devices
• Provide long-term storage of programs and data

5
Different types of Primary and Secondary Storage
Why do we need so many different kinds of storage
devices?
6
Storage Hierarchy
7
Five Characteristics of Storage Devices

• Speed
• Volatility
• Access method
• Portability
• Cost and capacity

8
Characteristic 1 Speed

• Access time
• the time required to execute one complete
  read/write operation
• For some devices access time is constant
  regardless storage location. (e.g. RAM)
• For others access time varies with storage
  location. (e.g. HD)
• Average access time
• Primary storage speed
• Typically faster than secondary storage speed by
  a factor of 105 or more
• measured in nanoseconds (1 ns 1 billionths of
  a second)
• Very important to overall system performance
• Recall wait state
• Secondary storage speed
• measured in milliseconds (1 ms 1 thousandths
  of a second)
• Important to some applications

9
Characteristic 1 Speed (2)

• Storage device speed is decided by
• Access time
• The unit of data transfer to/from the storage
  device
• For primary storage usually a word
• For secondary storage a block (512 bytes is
  typical block size)
• Data transfer rate 1 / access time x unit of
  data transfer
• e.g. a primary storage device with 15 ns access
  time, and word size is 4 bytes ? data transfer
  rate ?
• e.g. a typical hard disk with 50 ms access
  time, and block size is 512 bytes ? data transfer
  rate ?

10
Characteristic 2 Volatility

• Volatility is a matter of degree and conditions.
  Why?
• Primary storage devices are generally volatile
• Cannot reliably hold data for long periods
• Secondary storage devices are generally
  nonvolatile
• Hold data without loss over long periods of time

11
Characteristic 3 Access Method

• The physical structure determines the ways in
  which data can be accessed.
• Serial access (linear)
• Access time depends on the current position of
  read/write mechanism and the position of the
  desired data item.
• Usually hold backup copies of data, e.g. magnetic
  tape
• Random access (direct access)
• Is not restricted to any specific order when
  accessing data
• Access time may or may not be constant, e.g. RAM,
  HD
• Parallel access (simultaneous)
• Can simultaneously access multiple storage
  locations
• e.g. RAM, HD with some OS

12
Characteristic 4 Portability

• Removable storage media with standardized formats
  
• e.g., compact disc, tape, USB flash memory
• Typically results in slower access speeds
• Why?
• high-speed access requires tight control of
  environmental factor.
• e.g. In a HD, sealed enclosures minimize /
  eliminate dust and air density variations.

13
Characteristic 5 Cost and Capacity

• Cost increases
• With improved speed, volatility, or portability
• As access method changes
• serial ? random ? parallel access
• Compromise between cost and other characteristics
• Primary storage expensive, high speed and
  combination of parallel/random access methods
• Secondary storage less expensive, slower, and
  capacity is greater

14
Summary Characteristics Cost
15
Primary Storage Devices

• Critical performance characteristics of primary
  storage are
• Access speed
• Data transfer unit size
• Must closely match CPU speed and word size to
  avoid wait states

16
Storing Electrical Signals

• Digital electrical signals
• Data are represent as digital electrical signals
  inside CPU.
• Digital electrical signals are the basis of data
  transmission among of devices
• Can be stored directly or indirectly

17
Storing Electrical Signals Directly

• Directly storing electrical power
• by devices such as batteries and capacitors
• Trade off between access speed and volatility
• Battery slow to accept / regenerate electrical
  current, but stable
• Capacitor charge / discharge faster, but lose
  charge quickly ? need to recharge frequently

18
Storing Electrical Signals Indirectly

• Indirectly storing electrical power
• Uses energy to alter the state of a device, such
  as a mechanical switch, or a magnetic field
• Inverse the process regenerates equivalent
  electrical signal
• Modern computers use memory implemented with
  semiconductors (RAM and ROM)

19
Random Access Memory (RAM)

• Characteristics of RAM
• Microchip implementation using semiconductors
• To read and write with equal speed
• Random access to stored bytes, words, or larger
  data units
• Basic types
• Static RAM (SRAM) entirely uses transistors
• (see the flip-flop circuit on the next slide)
• Dynamic RAM (DRAM) uses transistors and
  capacitors

(Details on how SRAM and DRAM work are not
required.)
20

• A flip-flop circuit remembers its last position
  (0/1)
• Each flip-flop circuit stores one bit, with
  additional 2 - 4 transistors perform read/write

21
Dynamic RAM (DRAM)

• 1 transistors 1 capacitors per bit
• stores each bit of data in a separate capacitor.
• As real-world capacitors are not ideal and hence
  leak electrons, the information eventually fades
  unless the capacitor charge is refreshed
  periodically (thousands time / sec)
• Less complex than SRAM (6 transistors per bit)
• ? could have higher density than SRAM
• ? Less expensive than SRAM
• Slower than SRAM
• Required refresh cycles
• Less efficient circuitry for accessing individual
  bit
• (typical access time SRAM 5 ns vs. DRAM 50 ns.
  )
• Compare What is the cycle time for a CPU with 4
  GHz clock rate?

22
(No Transcript)
23
Improve RAM Performance

• To bridge the performance gap between memory
• and CPU, three technologies are used
• Read-ahead memory access
• Activating the read/write circuitry need extra
  time
• Programs usually access memory sequentially
• ? Activating the read/write circuitry for
  location n1 during or after an access to
  location n
• Synchronous read operations (SDRAM)
• Write/read operation are broken into steps
• ? pipelining multiple write/read operations
• On-chip memory caches
• Enhanced DRAM (EDRAM) Puts a small amount of
  SRAM in DRAM, as cache

24
Nonvolatile Memory (NVM)

• NVM random access memory with long-term or
  permanent data retention
• e.g. NVM is used to store system BIOS, store
  programs and data in portable devices (handheld
  computers and cell phones )
• Firmware software resides in NVM
• NVM is slower than RAM
• Three generations of NVM
• ROM the content is permanent put into it
• EPROM (Erasable programmable ROM)
• EEPROM (Electronically Erasable programmable ROM)
• Why do not use NVM instead of volatile memory?

25
Some Common NVM In Use

• Flash RAM
• Competitive with DRAM in capacity and read
  performance
• Relatively slow write speed
• Limited number of write cycles (wear out)
• Primary used for secondary storage and for
  firmware that isnt frequently updated.
• Some other NVM technologies are under
  development Ferroelectric RAM, Polymer memory,
  (not required)

26
Memory Packaging

• Memory circuits are embedded in chips ? groups of
  chips
• are packed on a small circuit board
• Dual in-line packages (DIPs)
• Early RAM and ROM circuits
• Single in-line memory module (SIMM)
• Standard RAM package in late 1980s
• Double in-line memory module (DIMM)
• Newer packaging standard

27
CPU Memory Access

• Physical organization of memory a sequence of
  contiguous memory cells
• Big endian vs. little endian
• Computer manufactures made different design
  decision
• Big endian the most significant byte at the
  lowest memory address
• Little endian the other way around
• example
• Addressable memory the highest numbered storage
  byte can be represent -- determined by the
  number of bits used to represent an address
• e.g. 32-bit used to represent address

28
Memory Allocation and Addressing

• Memory allocation
• Assignment of specific memory addresses to
  system software, application programs, and data

29
Memory Allocation Schemes

• Absolute addressing
• memory address operands that refers to actual
  physical memory address
• (Whatre the disadvantages of absolute
  addressing?)
• Indirect addressing (a.k.a. relative
  addressing)
• Each program are written as though the first
  instruction is at address 0.
• CPU converts this relative address into physical
  address through the program offset ? offset
  register holds the offset value.

30
Memory Allocation for Multiple Programs
31
Magnetic Storage

• Physics the duality of magnetism and
  electricity electromagnetism is the physics of
  the electromagnetic field
• Converts electrical signals into magnetic charges
• Captures magnetic charge on a storage medium
• Polarity of magnetic charge represents bit
  values zero and one
• Later regenerates electrical current from stored
  magnetic charge

32
Principles of Magnetic Storage

• Components
• Write operation
• Read operation

33
Characteristics of Magnetic Storage
X
X
34
Coercivity and Areal Density

• Coercivity the ability of a substance or
  magnetic storage medium to accept and hold
  magnetic charges.
• Areal density a function of the length and width
  of an individual bit area.

35
Magnetic Tape

• Ribbon of plastic with a coercible (usually
  metallic oxide) surface coating
• Mounts in a tape drive for reading and writing
• Relatively slow serial access
• Compounds magnetic leakage wraps upon itself
• Susceptible to stretching, friction, temperature
  variations

36
Magnetic Tape

• Two approaches to recording data (no details
  needed)
• (a) Linear recording, (b) Helical scanning
• Several formats and standards (e.g., DDS DAT,
  AIT, Mammoth, DLT, LTO)

37
Magnetic Disk

• Flat, circular platter with metallic coating that
  is rotated beneath read/write heads
• Random access device read/write head can be
  moved to any location on the platter
• Hard disks floppy disks
• Cost performance leader for general-purposeon-lin
  e secondary storage

38
Components of a Disk Drive
39
Tracks, Sectors and Cylinder
40
To increase capacity per platter, disk
manufacturers divide tracks into zones and vary
the sectors per track in each zone.
41
Magnetic Disk Access Time

• Disk Access Steps
• Switch among read/write heads
• Position the heads over a track
• Wait for the desired sector to rotate beneath the
  heads
• Disk Delay
• Head-to-head switching time HTH (All heads share
  on set of circuit.)
• Track-to-track seek time TTT
• Rotational delay

42
Most important performance numbers

• Average access delay (e.g. p196)
• For a large number of random accesses, the
  expected HTH switch time is the switching time of
  half of the number of recording surfaces.
• The expected TTT seek time is the movement time
  over half of the tracks.
• The expected rotation delay is time needed to
  rotate half of a track.
• Average access time
• Average access time
• average access delay the time reading a
  sector
• Sequential access time the time reading a
  sector
• How to minimize average access time ?
• Organize related data in sequential sectors of
  the same a track
• Equivalently positioned tracks on multiple
  platters
• De-fragmentation

43
Optical Mass Storage Devices

• Store bit values as variations in light
  reflection
• Higher areal density and longer data life than
  magnetic storage (why?)
• The basic principle reflecting a laser off of a
  recording surface and detecting changes in the
  reflected light compared to the original light.
• Standardized and relatively inexpensive
• Uses low performance requirements, high capacity
  requirements, portable and in standardized format

44

• Optical storage devices read data by shining
  laser beam on the disc.
• Reflecting a laser off of a recording surface
  and detecting changes in the reflected light
  compared to the original light.
• Photoelectric cell is positioned at a
  complementary angle to intercept reflected laser
  light.

45
CD-ROM

• Read-only data permanently embedded in durable
  polycarbonate disc
• Bit values represented as flat areas (lands) and
  concave dents (pits) in the reflective layer
• Data recorded in single continuous track that
  spirals outward from center of disc
• Popular medium for distributing software and
  large data sets

46
CD-R and Magneto-Optical

• CD-R
• Uses a laser that can be switched between high
  and low power and a laser-sensitive dye embedded
  in the disc
• Relatively cheap
• Common uses create music CDs on home computers,
  back up data from other storage devices, create
  archives of large data sets, and manufacture
  small quantities of identical CDs
• Magneto-Optical
• Utilize both optical and magnetic technologies.
• Technology peaked in the mid 1990s. It is waning.

47
Phase-Change Optical Discs and DVD

• Phase-Change Optical Discs
• Enables nondestructive writing to optical storage
  media
• Materials change state easily from
  non-crystalline (amorphous), to crystalline, and
  then back again
• Example CD-RW
• DVD improves on CD and CD-RW technology
• Increased track and bit density smaller
  wavelength lasers and more precise mechanical
  control
• Improved error correction
• Multiple recording sites and layers

Read Business Focus on page 206. We will discuss
this topic after midterm.
48
Technologies and Storage formats for Optical
Storages