Systems Architecture, 7e Solutions-11

Ch. 5

Systems Architecture 7th Edition

Burd Solutions Manual
Full download at link:

Solution Manual: https://testbankpack.com/

Test Bank: https://testbankpack.com/

Chapter 5 Solutions
Vocabulary Exercises
1. Dynamic RAM requires frequent __________ to maintain its data content.
refresh cycles
2. The __________ rate is the speed at which data can be moved to or from a storage device over a
communication channel.
data transfer
3. Three standard optical storage media that are written only during manufacture are called
__________ and __________.
CD-ROM, DVD-ROM, BD
4. __________, __________, ___________, and __________ are competing standards for rewritable
DVD discs.
DVD-RW, DVD+RW, DVD+RW DL, and DVD-RW DL
5. The __________ of a hard disk drive generate or respond to a magnetic field.
read/write heads
6. Data stored on magnetic media for long periods of time might be lost because of _________ and
__________.
magnetic leakage, magnetic decay
7. The contents of most forms of RAM are __________, making them unsuitable for long-term data
storage.
volatile
8. __________ and __________ are promising new technologies for implementing NVM that
promise significant increases in longevity.
Systems Architecture, 7e Solutions-22
Ch. 5

magnetoresistive RAM (MRAM), phase-change memory (PCM)

9. __________ is a flash RAM technology that provides higher capacity than traditional
2D/planar flash RAM technology. Add answer
vertical flash RAM.
10. __________ is typically stated in milliseconds for secondary storage devices and nanoseconds for
primary storage devices.
Access time
11. The three components that are summed to calculate average access time for a disk drive are
__________, __________, and __________.
head-to-head switching time, track-to-track seek time, rotational delay
12. In a magnetic or optical disk drive, a read/write head is mounted on the end of a(n) __________.
access arm
13. A(n) __________ mimics the behavior and physical size of a magnetic disk drive but has no
moving parts.
solid-state drive
14. A(n) __________ is a series of sectors stored along one concentric circle on a platter.
track
15. The data transfer rate of a magnetic or optical disk drive can be calculated by dividing 1 by the
drive’s access time and multiplying the result by the __________.
data transfer unit size
16. __________, __________, and __________ are storage formats originally designed for music or
video recording that have been applied to computer data storage.
CD-DA, DVD, BD
17. Average access time can usually be improved by __________ files stored on a disk.
defragmenting
18. Modern desktop and laptop computers generally use memory packaged on small standardized
circuit boards called __________.
SIMMs or DIMMs
19. The __________ of a magnetic or optical storage medium is the ratio of bits stored to a unit of the
medium’s surface area.
areal density (or recording density or bit density)
20. For most disk drives, the unit of data access and transfer is a(n) __________ or __________.
block, sector
21. Software programs stored permanently in ROM are called __________.
firmware

Review Questions
1. What factors limit the speed of an electrically based processing device?
The most important factor is the device size and, therefore, the distance electricity must
travel to enable the device to perform its function. Another factor is resistance of the
Systems Architecture, 7e Solutions-33
Ch. 5

internal conductive pathways, which is a function of the materials composing these

pathways and the device’s operating temperature. Packing density, the type of packaging,
and the presence or absence of heat dissipation mechanisms influence operating
temperature, which influences resistance and reliability.
2. What are the differences between static and dynamic RAM?
The most important differences are speed and cost. Static RAM is 5 to 10 times faster and
approximately 10 times more expensive than dynamic RAM. Static RAM is composed
entirely of transistors, and dynamic RAM is composed of transistors and capacitors.
Capacitors require frequent recharging, which slows access time.
3. What improvements are offered by synchronous DRAM compared with ordinary DRAM?
Synchronous DRAM begins the next sequential read access on each clock cycle and
overlaps processing stages of several read accesses. Ordinary DRAM accepts and
processes a read request in its entirety before accepting and processing another read
request. Synchronous DRAM is several times faster than ordinary DRAM when
processing sequential read accesses.
4. Why isn’t flash RAM commonly used to implement primary storage?
Flash RAM wears out after several hundred thousand read/write cycles. Therefore, it
would have to be replaced frequently if used as primary storage in a general-purpose
computer.
5. Describe current and emerging nonvolatile RAM technologies. What potential advantages do the
emerging technologies offer compared with current flash RAM technology?
NVM technologies include ROM, EPROM, EEPROM, flash RAM, magnetoresistive
RAM (MRAM), and phase-change memory (PCM). Flash RAM is currently the
dominant NVM type. MRAM and PCM are emerging NVM types. The main advantage
of MRAM is that it doesn’t degrade with repeated writes, which gives it better longevity
than conventional flash RAM. The main advantages of PCM are faster write times and
better longevity than flash RAM. However, PCM has slower read times and lower storage
density than flash RAM.
6. Describe serial, random, and parallel access. What types of storage devices use each method?
Serial access reads or writes data units in sequential order. Magnetic tape is the only
widely used form of serial access storage. Random access can “jump” directly between
two noncontiguous data units. All primary storage and disk storage devices use random
access. Parallel access reads or writes portions of a data item in parallel on separate
storage devices or media. RAM can be considered a parallel access device.
7. How is data stored and retrieved on a magnetic mass storage device?
For writing, electrical current is routed through a read/write head, which generates a
magnetic charge at its tip. A bit area of a magnetic recording surface is placed in close
proximity to the read/write head, and a charge is induced in the surface material. The
direction of current flow through the read/write head determines the stored charge’s
polarity.
For reading, a bit area of the recording surface is placed in close proximity to the
read/write head, and the stored charge induces current to flow through the read/write
head. A switch connected to the read/write head detects the current flow’s direction to
interpret it as a bit value.
8. Describe the factors that contribute to a disk drive’s average access time. Which of these factors is
improved if spin rate is increased? Which is improved if areal density is increased?
Systems Architecture, 7e Solutions-44
Ch. 5

Disk average access time is a combination of rotational delay, track-to-track seek time
(head movement), and head-to-head switching time. Rotational delay is reduced if spin
rate is increased.
If areal density is increased, TTT seek time is improved because tracks are narrower and
packed together more tightly (more tracks per same unit of distance traveled by the
read/write head).
9. Which factors do magnetic storage device manufacturers balance to ensure high capacity and
performance at reasonable cost and reliability?

The primary balance is between coercivity of the media coating and the areal density because
increasing areal density increases volatility thus reducing reliability. Manufacturers also balance
spin rate with device life. Faster spin rates reduce rotational delay and thus improve performance.
But they also cause mechanical components such as bearings to wear/fail more quickly. Both
“balancing acts” can be shifted in favor of greater performance/lifetime by using more expensive
components such as better bearings and higher coercivity coatings though doing so increases cost
– yet another aspect of the manufacturer’s balancing act.

10. Why is the areal density of optical discs higher than the areal density of magnetic disks? What
factors limit this areal density?
A laser can be tightly focused to a narrow beam. Magnetic charge can’t be focused easily,
and an attempt to charge a very small area overwrites surrounding areas. Optical bit size
is limited by the read laser’s wavelength and the amount of reflected light required to
activate the read mechanism.
11. Describe the processes of reading from and writing to a phase-change optical disc. How does the
performance and areal density of these discs compare with magnetic disks?
A phase-change disc is coated with a material that’s amorphous or crystalline. Each state
has different reflective properties. A bit area in the amorphous state is changed to a
crystalline state by heating it to a precise temperature with a laser. A bit area in the
crystalline state is changed to the amorphous state by heating it to the material’s melting
point. Phase-change discs are read in the same fashion as other optical discs.
Phase-change optical discs have slower access time than magnetic disks because more
time is required to heat a bit with a laser than to alter its magnetic polarity. Phase-change
discs also lose their capability to change state easily after repeated use and, therefore,
have shorter lives than magnetic disk platters under continuous use.
12. List and briefly describe the standards for recordable and rewritable CDs and DVDs. Are any of
the standards clearly superior to their competitors?
This answer is taken from Table 5.4:

Technology/for Writable? Description

mat
CD-ROM No Adaptation of musical CD technology; 650 or 700 MB capacity.
CD-R One time only CD-ROM format with a dye reflective layer that can be written by a
low-power laser.
CD-RW Yes CD-ROM format with phase-change reflective layer; can be written up
to 1000 times.
DVD-ROM No Adaptation of DVD video technology; similar to CD-ROM but more
advanced; 4.7 GB (single layer) or 8.5 GB (dual layer) capacity.
DVD+/-R One time only DVD-ROM single- and dual-layer formats; similar to CD-R with
improved performance and capacity. DVD-R and DVD+R are slightly
different formats
Systems Architecture, 7e Solutions-55
Ch. 5

DVD+/-RW Yes DVD-ROM single- and dual-layer formats with phase-change

reflective layer. DVD-RW and DVD+RW are slightly different
formats.

13. Why should you defragment magnetic disk drives? Why should you avoid defragmenting SSDs?

Defragmenting a disk drive reorganizes physical storage so that sectors of the same file are stored
in a way that minimizes rotational delay and head movement during sequential read operations. In
most cases, this significant improves read performance.

Although SSDs emulate the sector-oriented architecture of a disk drive, they use a fundamentally
different physical architecture that imposes no read performance penalty for accessing sectors
scattered throughout the device. Thus, defragmentation provides no performance improvement.
Further, SSDs have a limited number of write cycles. Defragmentation, which consumes many
write cycles over many storage locations, contributes to premature drive failure.

14. In what way(s) is/are SSDs an improvement over magnetic disk drives? In what way(s) isn't/aren't
they an improvement over magnetic disk drives?

By avoiding all mechanical delays in accessing storage locations, SSDs yield a significant
performance improvement compared to magnetic disk drives. However, their lifetime, as
measured in write cycles, is significantly less than magnetic disk drives. Also, their cost per unit
of storage is generally higher than for magnetic disk drives.

Problems and Exercises

1.
a. Assuming a sector size of 512 bytes, the capacity is 5 platters ×2 sides ×1024 tracks ×50
sectors ×512 bytes, or 262,144,000 bytes (250 MB).

b. Assuming the data is organized for maximum read efficiency, the read occurs cylinder by
cylinder. This results in 1024 track-to-track seeks (1024 ×3 microseconds = 0.003072
seconds) and 10 head-to-head switches per track (1024 ×10 × 2 nanoseconds = 0.00002048
seconds). The number of disk rotations equals the number of tracks read (1024 × 10 = 10,240
tracks). The disk rotates at 10,000 rpm, so the time required to read the tracks (exclusive of
seeks and switches) is 10,240 ÷ 10,000, or 1.024 minutes (61.44 seconds). Because there’s no
rotational delay when reading sequentially, the total elapsed time to read the disk is 0.003072
+ 0.00002048 + 61.44 = 61.44309248 seconds.

c. Serial access time is the time required to read the second of two sequential sectors. If the
sectors are on the same platter and track, it’s only a function of rotation speed. Because there
are 50 sectors per track, the sequential access time is 1/50 of a rotation, or 0.00012 seconds
(60 seconds ÷ 10,000 ÷ 50). An average serial access time should also account for the
fraction of sequential reads that require a head-to head switch (1 every 50 sectors) and a track-
to-track seek (1 every 500 sectors). The easiest way to compute this adjusted access time is to
divide the time required to read the entire disk by the number of sectors (61.44309248
seconds ÷ 512,000 sectors = 0.00012000604 seconds).

d. Each read requires one-half rotational delay, a seek over 512 tracks, switching through 5
heads, and 1/50 of a rotation to read the data. One-half rotational delay is 60 seconds ÷ 10,000
÷ 2, or 0.003 seconds. A seek over 512 tracks requires 512 × 0.000003, or 0.001536 seconds.
Switching through 5 heads requires 5 × 0.000000002, or 0.00000001 seconds. Reading the
Systems Architecture, 7e Solutions-66
Ch. 5

data requires 0.00012 seconds (the unadjusted serial access time). The average access time is
the sum of these numbers, or 0.00465601 seconds (4.65 milliseconds).
2.
a. 1 ÷ 2,400,000,000 ÷ 2 = approximately 0.2083 nanoseconds. Note that this answer allows no
time for transmitting the access request.

b. Processor cycle time is 0.4167 nanoseconds, so 12 CPU cycles (5 divided by 0.4167) are
required to complete a fetch operation. Half of the next cycle (the fetch portion) is wasted
because execution can’t begin until 16.25 seconds have elapsed. Therefore, six wait states are
incurred.

c. 10 ÷ 0.4167 = 24 CPU cycles to complete the fetch. Part of the next cycle is wasted waiting
for the execution cycle to begin, so 24 wait states are incurred.

3.

Storage device Average access time Data transfer unit size Data transfer rate

RAM 4 ns 64 bits (1 ÷ .000000004 seconds) ×

64 bits = 16 Gbps (or 2
GBps)
Optical disc 100 ms 512 bytes (1 ÷ .01 seconds) × 512
bytes = 5,120 Bps (or 5.12
KBps)
Note: This computation is a
worst-case scenario. The
data transfer rate is much
higher for sequential access.
Magnetic disk 5 ms 512 bytes (1 ÷ .005 seconds) × 512
bytes = 102,400 Bps (or
102.4 KBps)
Note: This computation is a
worst-case scenario. The
data transfer rate is much
higher for sequential access.

Research Problems

Project 1

Answers will vary based on the computers selected for the investigation, but students
might note some of these general findings: Older computers might not be able to accept a
1 GB memory expansion. The cost of 1 GB of memory for most current desktops and
laptops is similar. The cost of 1 GB of memory for most servers is higher because
memory must often be added in groups of four.
Project 2

SSDs should have uniformly higher performance than magnetic drives. But SSD cost per
GB should be higher and multi-terabyte drives should be few and far between in 2015 or
2016. Hybrid drives should provide nearly the same performance characteristics as SSDs
but at reduced cost and with higher capacities.
Project 3
Systems Architecture, 7e Solutions-77
Ch. 5

Answers will vary based on students’ opinions, but one comparison outcome (at the time
of this writing - 2015) is as follows:
• CD+R stores 700 MB and costs less than 25¢ each if bought in bulk.
• CD+RW also stores 700 MB but can be rewritten; costs about $1.25 each.
• DVD+R and DVD-R store 4.7 GB and cost about 25¢ each if bought in bulk.
• DVD+RW and DVD-RW also store 4.7 GB but cost about $1.25 each.
• BD-RE stores 25 GB and costs about $2.50 each if bought in bulk.
• The functional lifetime of all recordable formats is similar.
• All rewritable media are prone to data degradation if data is rewritten repeatedly
over time.
Storing 20 GB of digital photographs would take 29 CD-Rs (at a cost of about $7.25), or
6 DVD-Rs (at a cost of about $3.75), or 1 BD-RE (at a cost of about $2.50). BD-RE
appears to be the preferred solution based on lower cost and more compact storage.
High-capacity flash drives and cloud-based storage are attractive alternatives to
recordable optical media. Flash drive access time is much faster than optical media,
especially for writes. As of 2015, cost per GB is about 25¢ per GB up to 128 GB – about
2.5X the cost of BD-RE media but bypassing the cost of a compatible drive.
Cloud-based storage is relatively cheap and it removes the purchaser from worries about
long-term device lifetime and compatibility. But access times are limited by network
connection speeds which can be very slow in many home environments. Vendor
reliability and data privacy must also be considered.