Q.1 Which of the following is a common metric used to evaluate CPU performance?
Power consumption
Clock speed
Instruction per cycle (IPC)
Cache size
Explanation - IPC measures how many instructions a CPU can execute per clock cycle, which is a direct indicator of performance.
Correct answer is: Instruction per cycle (IPC)
Q.2 If a processor executes 10^9 instructions in 2 seconds, what is its average instruction execution rate?
0.5 GHz
0.5 GIPS
2 GHz
2 GIPS
Explanation - GIPS stands for Giga Instructions Per Second. 10^9 instructions / 2 seconds = 0.5 GIPS.
Correct answer is: 0.5 GIPS
Q.3 Amdahl's Law is used to:
Calculate CPU power consumption
Determine maximum speedup with parallelization
Measure cache latency
Compute instruction set size
Explanation - Amdahl's Law estimates the theoretical maximum speedup of a task using parallel processors, considering the serial portion.
Correct answer is: Determine maximum speedup with parallelization
Q.4 Which factor has the most direct impact on dynamic power consumption in CMOS circuits?
Clock frequency
Cache size
Instruction set
Number of registers
Explanation - Dynamic power is proportional to the switching activity, capacitance, voltage squared, and clock frequency.
Correct answer is: Clock frequency
Q.5 Which metric measures energy efficiency as performance per watt?
MIPS
FLOPS
Performance per Watt
IPC
Explanation - Performance per Watt evaluates how efficiently a system converts power into computational performance.
Correct answer is: Performance per Watt
Q.6 CPI stands for:
Cycles per Instruction
Cache Performance Index
Clock Power Indicator
Computational Performance Index
Explanation - CPI measures the average number of clock cycles each instruction takes to execute.
Correct answer is: Cycles per Instruction
Q.7 Which of the following reduces power consumption without affecting performance significantly?
Increasing clock frequency
Voltage scaling
Adding more cores
Increasing cache size
Explanation - Reducing voltage decreases dynamic power quadratically while maintaining the same clock frequency if timing allows.
Correct answer is: Voltage scaling
Q.8 What is the main goal of the SPEC benchmark suite?
Measure energy efficiency
Evaluate processor performance
Estimate chip area
Determine manufacturing cost
Explanation - SPEC (Standard Performance Evaluation Corporation) benchmarks are used to assess CPU performance with standardized workloads.
Correct answer is: Evaluate processor performance
Q.9 In CPU performance analysis, execution time is given by:
Instruction count × CPI × Clock cycle time
Instruction count × Clock frequency
CPI × Voltage
Instruction count ÷ Clock frequency
Explanation - Execution time = Number of instructions × Cycles per instruction × Time per cycle.
Correct answer is: Instruction count × CPI × Clock cycle time
Q.10 What does the term 'thermal design power (TDP)' represent?
Maximum performance under load
Maximum heat a CPU generates under typical load
Instruction execution time
Number of transistors
Explanation - TDP is the estimated maximum power a cooling system needs to dissipate under typical workloads.
Correct answer is: Maximum heat a CPU generates under typical load
Q.11 Which of the following reduces dynamic power in processors?
Clock gating
Increasing pipeline depth
Adding more registers
Instruction reordering
Explanation - Clock gating turns off the clock to idle portions of a circuit, reducing unnecessary switching and dynamic power consumption.
Correct answer is: Clock gating
Q.12 The relationship between power, voltage, and frequency in CMOS circuits is approximately:
P ∝ V^2 × f
P ∝ f/V
P ∝ V/f
P ∝ V × f^2
Explanation - Dynamic power in CMOS is proportional to the capacitance switched per cycle, voltage squared, and clock frequency.
Correct answer is: P ∝ V^2 × f
Q.13 Which metric is often used to compare the energy efficiency of supercomputers?
LINPACK GFLOPS/Watt
MIPS
Cache hit ratio
Instruction latency
Explanation - Supercomputer energy efficiency is often measured as floating-point operations per second per watt using LINPACK benchmarks.
Correct answer is: LINPACK GFLOPS/Watt
Q.14 If two CPUs have the same execution time but different power consumption, which is more power efficient?
The CPU with higher power
The CPU with lower power
Both are equal
Cannot determine from given data
Explanation - Lower power consumption for the same performance indicates higher energy efficiency.
Correct answer is: The CPU with lower power
Q.15 Which factor does NOT directly affect CPU performance?
Instruction count
Cache size
CPU color
CPI
Explanation - CPU color is purely cosmetic and has no impact on execution speed.
Correct answer is: CPU color
Q.16 Which type of power is consumed even when a CPU is not switching any transistors?
Dynamic power
Leakage power
Instruction power
Clock power
Explanation - Leakage power is due to small currents in transistors even when they are idle.
Correct answer is: Leakage power
Q.17 Which of the following improves performance per watt?
Increasing frequency indefinitely
Adding more efficient cores
Disabling caches
Increasing voltage unnecessarily
Explanation - Efficient cores provide better performance with less energy per computation, improving performance per watt.
Correct answer is: Adding more efficient cores
Q.18 Which of these is an example of a performance benchmark?
SPEC CPU
TDP rating
CPI value
Voltage rating
Explanation - SPEC CPU is a widely used benchmark suite to evaluate processor performance using standardized workloads.
Correct answer is: SPEC CPU
Q.19 The formula for CPU execution time can also be written as:
CPU time = Instruction count / (CPI × Clock frequency)
CPU time = Instruction count × CPI / Clock frequency
CPU time = CPI × Clock frequency × Instruction count
CPU time = Clock frequency / (Instruction count × CPI)
Explanation - Execution time can be expressed as the total cycles (Instruction count × CPI) divided by the clock frequency.
Correct answer is: CPU time = Instruction count × CPI / Clock frequency
Q.20 Which is true about multicore processors and energy efficiency?
Multiple cores always consume more power than a single core
Multiple cores can improve energy efficiency by parallel execution at lower frequency
Multiple cores reduce performance
Energy efficiency is unrelated to core count
Explanation - By distributing workload across multiple cores running at lower frequency, energy per operation can decrease.
Correct answer is: Multiple cores can improve energy efficiency by parallel execution at lower frequency
Q.21 Which of the following is used to measure how much faster a system executes a task compared to a reference?
Speedup
CPI
TDP
Voltage scaling
Explanation - Speedup is the ratio of execution time on a reference system to execution time on the target system.
Correct answer is: Speedup
Q.22 Dynamic Voltage and Frequency Scaling (DVFS) is used to:
Increase transistor count
Reduce power while maintaining performance
Increase instruction count
Measure CPI
Explanation - DVFS adjusts voltage and frequency according to workload to save power without significantly reducing performance.
Correct answer is: Reduce power while maintaining performance
Q.23 Which component typically dominates total CPU power consumption at high frequencies?
Leakage power
Dynamic power
Cache power
I/O power
Explanation - At high clock frequencies, dynamic power (switching transistors) dominates over leakage power.
Correct answer is: Dynamic power
Q.24 Which metric considers both execution time and power consumption?
Energy-delay product
CPI
Clock frequency
Instruction count
Explanation - Energy-delay product combines energy consumed and execution time to evaluate overall efficiency.
Correct answer is: Energy-delay product