Higher School of Sciences and 2nd cycle: 1st class

Technologies of Computing and Digital Year: 2023/2024

(ESTIN) Semestre : 2

Recitation 1 (Part 3 and final):

Comprehensive Performance Assessment Across Architectures

Exercise 10:
The modern microprocessors can adjustable their voltage. If we suppose that a 15% reduction in voltage may
result in a 15% reduction in frequency. What would be the impact on dynamic energy and on dynamic power?

Exercise 11:
You are designing a system for a real-time application in which specific deadlines must be met. Finishing the
computation faster gains nothing. Completing the computation faster provides no advantage. It is observed
that the system can execute the required code at a rate twice as fast as needed in the worst case.

1. Why is there no advantage in finishing the computation faster?

2. How much energy can be saved by maintaining the current speed and powering off the system once
the computation is done?

3. How much energy do you save if you set the voltage and frequency to be half as much?

Problem 1

We want to improve the performance of a processor with a clock frequency of 1 GHz and which has the
characteristics listed in the table below:

Instruction type / Functional unit CPI

Arithmetic and logic 2
Integer multiplication 5
Floating point operation 18
Branch 5
Load/store 8

In order to assess the processor performance, you are given three typical programs that you will use for
testing. These programs have the following characteristics in terms of instructions count (in millions):

Program 1 Program 2 Program 2

Arithmetic and logic 150 55 30
Integer multiplication 25 0 20
Floating point operation 0 125 300
Branch 80 30 20
Load/store 70 10 60

1. Calculate the throughput of the three programs in MIPS and MFLOPS.

2. Calculate the execution time of the three programs.
3. After working on the optimization of the 5 units of the processor, you propose three improvements as
listed below.

• Improvement 1: The execution time of floating-point operations can be accelerated by a

factor of 3.

• Improvement 2: The execution time of branches can be accelerated by a factor of 2.

• Improvement 3: The execution time of memory accesses can be accelerated by a factor of 2.

Using the three programs for testing, which improvement provides the best acceleration of the
processor? Discuss and justify the answer.

Page 1 sur 2
 Higher School of Sciences and 2nd cycle: 1st class
Technologies of Computing and Digital Year: 2023/2024
(ESTIN) Semestre : 2

Problem 2

A cell phone performs very different tasks, including streaming music, streaming video, and reading email.
These tasks perform very different computing tasks. Battery life and overheating are two common problems
for cell phones, so reducing power and energy consumption are critical.
In this problem, we consider what to do when the user is not using the phone to its full computing capacity.
For these problems, we will evaluate an unrealistic scenario in which the cell phone has no specialized
processing units. Instead, it has a quad-core, general purpose processing unit. Each core uses 0.5 W at full
use. For email-related tasks, the quad-core is 8 times as fast as necessary.
1. How much dynamic energy and power are required compared to running at full power?
First, suppose that the quad-core operates for 1/8 of the time and is idle for the rest of the time. That
is, the clock is disabled for 7/8 of the time, with no leakage occurring during that time. Compare total
dynamic energy as well as dynamic power while the core is running.

2. How much dynamic energy and power are required using frequency and voltage scaling? Assume
frequency and voltage are both reduced to 1/8 the entire time.

3. Now assume the voltage may not decrease below 50% of the original voltage. This voltage is referred
to as the voltage floor, and any voltage lower than that will lose the state. Therefore, while the
frequency can keep decreasing, the voltage cannot. What are the dynamic energy and power savings
in this case?

Page 2 sur 2