0% found this document useful (0 votes)
37 views3 pages

Assignment 6

The document contains 10 multiple choice questions and 2 short answer questions about embedded systems, parallel processing, and computer architecture topics like VLIW, dark silicon, and ARM Thumb instruction set. The questions test knowledge of concepts like energy efficiency techniques in embedded systems, advantages of different chip designs like ASIC and FPGA, definitions of dark silicon and assumptions of scheduling algorithms, properties of parallel processing systems and VLIW architecture, and basics of ARM Thumb instructions.

Uploaded by

Roopa Patil
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
37 views3 pages

Assignment 6

The document contains 10 multiple choice questions and 2 short answer questions about embedded systems, parallel processing, and computer architecture topics like VLIW, dark silicon, and ARM Thumb instruction set. The questions test knowledge of concepts like energy efficiency techniques in embedded systems, advantages of different chip designs like ASIC and FPGA, definitions of dark silicon and assumptions of scheduling algorithms, properties of parallel processing systems and VLIW architecture, and basics of ARM Thumb instructions.

Uploaded by

Roopa Patil
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

Assignment‌‌6‌ ‌

MCQ:‌ ‌
1.‌‌To‌‌reduce‌‌the‌‌energy‌‌consumption‌‌of‌‌an‌‌embedded‌‌system,‌‌there‌‌needs‌‌to‌‌be‌‌at‌‌most‌‌____‌‌
discrete‌‌voltage‌‌levels.‌ ‌
A. Three‌ ‌
B. Four‌
C. Two‌ ‌
D. Five‌ ‌
Solution:‌‌Option‌‌C.‌‌Two‌ ‌
Explanation:‌‌This‌‌is‌‌the‌‌statement‌‌of‌‌the‌‌Ishihara‌‌Yasuura‌‌Lemma.‌ ‌

2.‌ ‌Suppose‌ ‌I‌ ‌have‌ ‌to‌‌build‌‌a‌‌device‌‌that‌‌has‌‌to‌‌perform‌‌multiplication‌‌of‌‌2-bit‌‌numbers.‌‌Which‌‌
chip‌‌design‌‌will‌‌let‌‌me‌‌perform‌‌multiple‌‌multiplications‌‌at‌‌highest‌‌energy‌‌efficiency?‌ ‌
A. DSP‌
B. ASIC‌ ‌
C. FPGA‌ ‌
D. None‌‌of‌‌the‌‌above‌ ‌
Solution:‌‌Option‌‌B.‌‌ASIC‌ ‌
Explanation:‌‌ASIC‌‌has‌‌the‌‌highest‌‌throughput‌‌in‌‌terms‌‌of‌‌operations‌‌per‌‌joule‌‌followed‌‌by‌‌FPGA‌‌
and‌‌DSP.‌ ‌

3.‌‌What‌‌is‌‌dark‌‌silicon‌‌with‌‌reference‌‌to‌‌the‌‌electronic‌‌industry?‌ ‌
A. Amount‌ ‌of‌ ‌circuitry‌ ‌on‌ ‌a‌ ‌chip‌ ‌which‌ ‌cannot‌ ‌be‌ ‌powered‌‌at‌‌the‌‌given‌‌operating‌‌voltage‌‌
due‌‌to‌‌thermal‌‌constraints‌‌of‌‌the‌‌chip.‌ ‌
B. The‌‌dark‌‌coloured‌‌silicon‌‌used‌‌in‌‌the‌‌transistors.‌ ‌
C. Both‌‌A‌‌and‌‌B.‌ ‌
D. None‌‌of‌‌the‌‌above.‌ ‌
Solution:‌ ‌Option‌ ‌A.‌ ‌Amount‌ ‌of‌ ‌circuitry‌ ‌on‌ ‌a‌ ‌chip‌ ‌which‌ ‌cannot‌ ‌be‌ ‌powered‌ ‌at‌ ‌the‌ ‌given‌‌
operating‌‌voltage‌‌due‌‌to‌‌thermal‌‌constraints‌‌of‌‌the‌‌chip.‌ ‌
Explanation:‌‌Option‌‌A‌‌is‌‌the‌‌definition‌‌of‌‌dark‌‌silicon.‌ ‌

4.‌‌Which‌‌are‌‌the‌‌assumptions‌‌in‌‌the‌‌SD‌‌algorithm?‌ ‌
A. Arrival‌‌and‌‌deadline‌‌time‌‌of‌‌all‌‌tasks‌‌are‌‌known.‌ ‌
B. Worst-case‌‌execution‌‌time‌‌is‌‌not‌‌known.‌ ‌
C. Both‌‌A‌‌and‌‌B.‌ ‌
D. None‌‌of‌‌the‌‌above.‌ ‌
Solution:‌‌Option‌‌A.‌‌Arrival‌‌and‌‌deadline‌‌time‌‌of‌‌all‌‌tasks‌‌are‌‌known.‌ ‌
Explanation:‌‌Worst-case‌‌execution‌‌time‌‌is‌‌known‌‌in‌‌case‌‌of‌‌the‌‌SD‌‌algorithm,‌‌hence‌‌option‌‌B‌‌is‌‌
incorrect.‌‌Option‌‌A‌‌is‌‌a‌‌fact.‌ ‌





5.‌‌In‌‌a‌‌system‌‌performing‌‌parallel‌‌processing‌‌of‌‌n‌‌tasks,‌‌the‌‌power‌‌of‌‌the‌‌system‌‌is‌ ‌
A. Directly‌‌proportional‌‌to‌‌n‌ ‌
B. Inversely‌‌proportional‌‌to‌‌n‌ ‌
C. Directly‌‌proportional‌‌to‌‌n2 ‌
D. Inversely‌‌proportional‌‌to‌‌n2 ‌
Solution:‌‌Option‌‌D.‌‌Inversely‌‌proportional‌‌to‌‌n2 .‌ ‌
Explanation:‌‌Ppar‌‌=‌‌Pseq/‌n2 .‌‌Therefore‌‌the‌‌power‌‌of‌‌a‌‌parallel‌‌processing‌‌system‌‌is‌‌inversely‌‌
proportional‌‌to‌‌n2 .‌ ‌

6.‌‌Which‌‌of‌‌the‌‌following‌‌properties‌‌about‌‌VLIW(Very‌‌Large‌‌Instruction‌‌Word)‌‌are‌‌false?‌ ‌
A. Reduces‌‌complexity‌‌of‌‌hardware‌ ‌
B. Reduces‌‌power‌‌consumption‌ ‌
C. Increases‌‌clock‌‌rate‌ ‌
D. Reduces‌‌program‌‌code‌‌size‌ ‌
Solution:‌‌Option‌‌D.‌‌Reduces‌‌program‌‌code‌‌size‌ ‌
Explanation:‌ ‌The‌ ‌program‌ ‌code‌ ‌size‌ ‌increases‌ ‌due‌ ‌to‌ ‌the‌ ‌multiple‌‌instructions‌‌and‌‌hence‌‌the‌‌
compilers‌‌and‌‌hardware‌‌is‌‌difficult‌‌to‌‌design.‌ ‌

7.‌‌Zero‌‌overhead‌‌loops‌ ‌
A. Automatically‌‌repeats‌‌the‌‌body‌‌of‌‌the‌‌loops.‌ ‌
B. Can‌‌be‌‌implemented‌‌both‌‌in‌‌hardware‌‌and‌‌software.‌ ‌
C. Speeds‌‌up‌‌the‌‌processing‌‌of‌‌the‌‌device.‌ ‌
D. All‌‌of‌‌the‌‌above.‌
Solution:‌‌Option‌‌D.‌‌All‌‌of‌‌the‌‌above‌ ‌
Explanation:‌ ‌Features‌ ‌of‌ ‌the‌ ‌Zero‌ ‌overhead‌ ‌loop‌ ‌instruction‌ ‌which‌ ‌is‌ ‌common‌ ‌in‌ ‌DSPs‌ ‌and‌‌
CISC‌‌instruction‌‌sets.‌ ‌

8.‌‌In‌‌a‌‌system‌‌performing‌‌parallel‌‌processing‌‌of‌‌n‌‌tasks,‌‌the‌‌time‌‌to‌‌run‌‌the‌‌program‌‌is‌‌____‌‌the‌‌
time‌‌taken‌‌by‌‌the‌‌system‌‌executing‌‌sequentially.‌ ‌
A. Same‌‌as‌ ‌
B. Greater‌‌than‌ ‌
C. Less‌‌than‌ ‌
D. Slightly‌‌less‌‌than‌ ‌
Solution:‌‌Option‌‌A.‌‌Same‌‌as‌ ‌
Explanation:‌‌The‌‌time‌‌is‌‌not‌‌affected‌‌by‌‌parallel‌‌scheduling‌‌of‌‌tasks.‌‌Only‌‌the‌‌power‌‌consumed‌‌
varies.‌ ‌

9.‌ ‌“VLIW‌ ‌is‌ ‌used‌ ‌in‌ ‌RISC‌ ‌architecture‌ ‌so‌ ‌that‌ ‌it‌ ‌can‌ ‌exploit‌ ‌instruction-level‌ ‌parallelism‌ ‌and‌‌
provide‌‌a‌‌cheaper‌‌and‌‌faster‌‌execution‌‌device.”‌ ‌
A. True‌ ‌
B. False‌ ‌
C. Cannot‌‌be‌‌determined‌ ‌
Solution:‌‌Option‌‌A.‌‌True‌ ‌
Explanation:‌‌VLIW‌‌stands‌‌for‌‌Very‌‌Long‌‌Instruction‌‌Words‌‌with‌‌parallel‌‌dispatch‌‌and‌‌is‌‌used‌‌to‌‌
schedule‌ ‌multiple‌ ‌instructions‌ ‌at‌ ‌once‌ ‌thus‌ ‌reducing‌ ‌the‌ ‌complicated‌ ‌nature‌ ‌of‌ ‌the‌ ‌used‌‌
hardware.‌ ‌

10.‌‌What‌‌is‌‌the‌‌bit‌‌size‌‌of‌‌the‌‌original‌‌ARM‌‌Thumb‌‌instruction‌‌set?‌ ‌
A. 64‌ ‌
B. 16‌ ‌
C. 32‌ ‌
D. None‌‌of‌‌the‌‌above‌ ‌
Solution:‌‌Option‌‌B.‌‌16‌ ‌
Explanation:‌‌Fact.‌‌The‌‌new‌‌Thumb2‌‌set‌‌has‌‌both‌‌16‌‌and‌‌32‌‌bit‌‌instructions.‌ ‌

Short-Answer‌‌type(Alphanumeric‌‌answers‌‌only):‌ ‌

11.‌‌Suppose‌‌a‌‌system‌‌with‌‌a‌‌clock‌‌of‌‌50Hz‌‌is‌‌made‌‌to‌‌run‌‌4‌‌tasks‌‌parallely.‌‌Calculate‌‌the‌‌new‌‌
clock‌‌frequency(upto‌‌two‌‌decimal‌‌places‌‌in‌‌Hz).‌ ‌
Solution:‌‌12.50‌ ‌

Explanation:‌‌fpar‌‌=‌‌fseq/n‌‌for‌‌n‌‌tasks‌‌run‌‌parallely.‌‌Putting‌‌fseq‌‌=‌‌50‌‌and‌‌n=4,‌‌fpar‌‌=‌‌12.50‌‌Hz‌ ‌

12.‌‌Suppose‌‌a‌‌system‌‌with‌‌a‌‌clock‌‌of‌‌50Hz‌‌runs‌‌4‌‌tasks.‌‌The‌‌power‌‌consumed‌‌for‌‌running‌‌all‌‌
the‌‌tasks‌‌sequentially‌‌is‌‌64W.‌‌Calculate‌‌the‌‌power‌‌to‌‌run‌‌all‌‌tasks‌‌per‌‌clock(in‌‌W).‌ ‌
Solution:‌‌16‌ ‌

Explanation:‌ ‌
Ppar‌‌=‌‌Pseq/‌n2 ‌and‌‌Ppar‌‌per‌‌clock‌‌=‌‌Ppar‌‌*‌‌n‌‌=‌‌Pseq/‌n .‌ ‌
Putting‌‌n‌‌=‌‌4‌‌and‌‌Pseq‌‌=‌‌64,‌‌Ppar‌‌per‌‌clock‌‌=‌‌16W.‌ ‌

You might also like