---
title: "Design goals and performance metrics"
subject: "Computer Organization and Architecture"
chapter: 1
key: "semester_5/computer_organization_and_architecture:design-goals-metrics"
source: https://www.bctengineer.com/q/semester_5/computer_organization_and_architecture/design-goals-metrics
site: BCT Engineering
---

# Design goals and performance metrics

Computer Organization and Architecture · Chapter 1: Introduction

- Asked in: 2076 Asoj, 5 marks
- Answer checked by Claude Opus 5.5 (high effort) on 2026-10-14
- Page: https://www.bctengineer.com/q/semester_5/computer_organization_and_architecture/design-goals-metrics
- Whole chapter: https://www.bctengineer.com/md/semester_5/computer_organization_and_architecture/chapter-1

## Question

Explain the design goals and performance metrics for a computer system regarding its organization and architecture.

## Short answer

Design goals are high performance, low cost, compatibility, reliability, scalability, and power efficiency. Performance metrics include clock speed, CPI, MIPS, CPU time, throughput, response time, speedup (Amdahl's law: 1/((1-f)+f/s)), and standard benchmarks.

## Answer

#### Design goals
The aim in designing computer organization and architecture is to get the best performance at a low cost with reliability and compatibility.
1. **Performance:** high speed and throughput.
2. **Cost:** low cost per unit of performance.
3. **Compatibility:** new machines should run old software (same instruction set).
4. **Reliability and availability:** the system should work correctly for long periods.
5. **Scalability and extensibility:** easy to add memory, devices and processors.
6. **Power efficiency and size:** especially for mobile and embedded systems.

#### Performance metrics
- **Clock speed:** number of clock cycles per second (Hz). Cycle time = 1 / f.
- **CPI (cycles per instruction):** average clock cycles needed for one instruction.
- **Instruction execution rate (MIPS) = f / (CPI × 10⁶).**
- **CPU time = (instruction count × CPI) / clock rate.**
- **MFLOPS:** million floating-point operations per second, used for scientific work.
- **Throughput:** number of tasks completed per unit time.
- **Response (execution) time:** time from start to finish of a task.
- **Speedup = time without improvement / time with improvement.** Amdahl's law: Speedup = 1 / ((1 - f) + f / s), where f is the fraction improved and s its speedup.
- **Benchmarks:** standard programs (for example SPEC) to compare machines fairly.

## Check yourself

The answer's key points as multiple-choice questions, with the correct option and why.

1. CPU time is given by
   A. clock rate x CPI
   B. instruction count / CPI
   C. (instruction count x CPI) / clock rate
   D. CPI / clock rate
   Answer: C. Total cycles divided by clock rate. (Key point: CPU time)

2. A 500 MHz processor has an average CPI of 2. What is its MIPS rate?
   A. 250
   B. 1000
   C. 500
   D. 2
   Answer: A. MIPS = f / (CPI x 10^6) = 500/2 = 250. (Key point: MIPS)

3. With fraction f = 0.5 improved by s = 2, what is the speedup by Amdahl's law?
   A. 2
   B. 1.5
   C. 1.33
   D. 0.75
   Answer: C. 1/((1-0.5)+0.5/2) = 1/0.75 = 1.33. (Key point: Amdahl's law)

4. What is throughput?
   A. Number of tasks completed per unit time
   B. Time to finish one task
   C. Cycles per instruction
   D. Clock speed
   Answer: A. Response time is the time for one task. (Key point: Throughput vs response)

## Related in Computer Organization and Architecture

- [Computer architecture vs organization](https://www.bctengineer.com/question.md?key=semester_5%2Fcomputer_organization_and_architecture%3Aarch-vs-org) — Computer Organization and Architecture, chapter 1, asked 5×
- [Performance balance and bus design elements](https://www.bctengineer.com/question.md?key=semester_5%2Fcomputer_organization_and_architecture%3Aperf-balance-bus-design) — Computer Organization and Architecture, chapter 1, asked 2×
- [Driving factors for performance design](https://www.bctengineer.com/question.md?key=semester_5%2Fcomputer_organization_and_architecture%3Aperf-design-factors) — Computer Organization and Architecture, chapter 1
- [Define computer architecture](https://www.bctengineer.com/question.md?key=semester_5%2Fcomputer_organization_and_architecture%3Adefine-arch) — Computer Organization and Architecture, chapter 1
- [Goals of plug and play BIOS](https://www.bctengineer.com/question.md?key=semester_5%2Fcomputer_organization_and_architecture%3Apnp-bios-goals) — Computer Organization and Architecture, chapter 7
- [How arithmetic pipeline increases performance](https://www.bctengineer.com/question.md?key=semester_5%2Fcomputer_organization_and_architecture%3Aarith-pipeline-performance) — Computer Organization and Architecture, chapter 4

## Related in other subjects

- [Design metrics of embedded system](https://www.bctengineer.com/question.md?key=semester_6%2Fembedded_system%3Adesign-metrics-embedded) — Embedded System, chapter 1, asked 3×
- [Goals and design challenges of DS](https://www.bctengineer.com/question.md?key=semester_7%2Fdistributed_system%3Ads-goals-challenges) — Distributed System, chapter 1, asked 3×
- [Design goals of digital modulation; coherent BPSK](https://www.bctengineer.com/question.md?key=semester_6%2Fcommunication_systems%3Adesign-goals-bpsk) — Communication Systems, chapter 7
- [Microprocessor based system and bus organization](https://www.bctengineer.com/question.md?key=semester_4%2Fmicroprocessor%3Amp-based-system-bus-organization) — Microprocessor, chapter 1, asked 5×