SILIGURI INSTITUTE OF TECHNOLOGY
NAME- SNEHA SARKAR
ROLL NO- 11900123134
DEPATMENT- CSE
SECTION- C
SEMESTER-Third
SUBMITTED TO-PROFESSEOR Mrs. AnkitaSinha
Introduction to Computer Architecture
• Definition:
• Computer architecture refers to the conceptual design and fundamental
operational structure of a computer system.
• It encompasses the hardware and software resources, as well as the logical
aspects of how data is processed and managed.
• Importance:
• Understanding computer architecture is crucial for optimizing performance,
enhancing computing efficiency, and guiding the development of new
technologies.
Basic Computer Architecture
• Components:
• Central Processing Unit (CPU): The brain of the computer, responsible for
executing instructions.
• Memory: Where data and instructions are stored temporarily or
permanently.
• Input/Output (I/O) Devices: Interfaces for the user and external devices to
interact with the computer.
• Bus: Communication pathways that transfer data between components.
• Process Flow:
• Instructions are fetched from memory, decoded by the CPU, executed, and
the results are stored back in memory or sent to an output device.
Harvard Architecture
• Definition:
• A computer architecture with physically separate storage and signal pathways for
instructions and data.
• Components:
• Separate Memories: Two distinct memory modules, one for instructions and another for
data.
• Separate Buses: Independent buses for transferring instructions and data, allowing
simultaneous access.
• Advantages:
• Increased Speed: Parallel processing of instructions and data enhances computational
speed.
• Efficiency: Minimizes bottlenecks by eliminating competition between instruction and data
fetches.
• Applications:
• Commonly used in embedded systems, digital signal processors (DSPs), and microcontrollers.
Von Neumann Architecture
• Definition:
• A computer architecture where both instructions and data share the same memory
space and data pathways.
• Components:
• Single Memory: Unified memory for storing both program instructions and data.
• Single Bus: A single bus system for fetching instructions and data sequentially.
• Advantages:
• Simplicity: Easier to design and program, making it suitable for general-purpose
computing.
• Flexibility: Can execute self-modifying code, useful for certain types of algorithms.
Von Neumann Bottleneck Problem
• Explanation:
• A performance limitation caused by the shared bus architecture, where the CPU must fetch
instructions and data sequentially, leading to delays.
• Causes:
• Shared Bus: The same bus is used for both instruction fetches and data transfers, leading to
competition and delays.
• Memory Access Delays: The CPU often waits for data or instructions to be fetched from
memory, slowing down execution.
• Impact:
• Reduced Throughput: Limits the maximum throughput of a system, affecting performance,
especially in high-speed computing.
• Mitigations:
• Caching: Utilizing small, fast memory caches to store frequently used data and instructions.
• Pipelining: Breaking down instruction execution into multiple stages that can be processed
concurrently.
Flynn’s Classification of Computer Architecture
• Overview:
• A classification system proposed by Michael J. Flynn in 1966, categorizing computer
architectures based on instruction and data streams.
• Categories:
• Single Instruction, Single Data (SISD):
• A traditional sequential architecture where one instruction operates on one data set at a time.
• Single Instruction, Multiple Data (SIMD):
• A parallel architecture where one instruction operates on multiple data sets simultaneously, often used in
vector processing.
• Multiple Instruction, Single Data (MISD):
• A rare and specialized architecture where multiple instructions operate on a single data set, often for
fault-tolerant systems.
• Multiple Instruction, Multiple Data (MIMD):
• A highly parallel architecture where multiple instructions operate on multiple data sets simultaneously,
used in multiprocessor systems.
• Applications:
• SISD in traditional CPUs, SIMD in GPUs, MISD in redundant systems, and MIMD in
supercomputers.
Conclusion
• Summary:
• Different computer architectures offer various strengths and weaknesses,
making them suitable for specific applications.
• The choice of architecture impacts performance, complexity, and power
efficiency.
• Future Trends:
• Ongoing research into quantum computing, neuromorphic architectures, and
advanced parallel processing techniques.
• Final Thoughts:
• Understanding these architectures helps in selecting the right system for
specific tasks and contributes to innovation in computing technologies.
