Computer Organization & Architecture Lecture 10

Continued from Computer Organization and Architecture Lecture 9

Computer Systems Organization & Architecture

The organization and architecture of a computer system are crucial in understanding how operating systems (OS) manage and allocate resources. These components must work seamlessly to ensure that the system functions efficiently. Below is a detailed yet concise overview of key elements that define a computer system’s organization and architecture.

Computer System Operations

Device Controllers

• What They Do: Device controllers manage hardware devices such as disk drives, audio devices, or video displays.
• Functionality: The CPU and device controllers can execute in parallel, competing for memory cycles.
• Importance: Efficient management of this parallel execution is essential to prevent performance bottlenecks and ensure that resources are used optimally.

Memory Controller

• What It Does: The memory controller synchronizes access to the system’s memory.
• Functionality: It coordinates the access requests from various components, ensuring that no conflicts occur when the CPU or device controllers attempt to read from or write to memory.
• Importance: Proper memory management is vital for maintaining system stability and performance.

Booting Process and Bootstrap Program

• What Happens During Booting: When the system is powered on, an initial program called the bootstrap program is loaded from ROM into EEPROM. This program is responsible for starting the system.
• Key Functions: The bootstrap program initializes system components like CPU registers and loads the OS kernel into memory. It must also know how to begin executing the operating system.
• Importance: Without a properly functioning bootstrap program, the system cannot load and execute the OS, halting all operations.

Kernel and System Processes

• Kernel’s Role: Once the kernel is loaded, it begins executing and providing core services to the system and users.
• System Daemons: These are system processes that are loaded at boot time and run alongside the kernel. In UNIX systems, for example, the first process is init, which starts other daemons.
• Importance: System daemons ensure the OS can provide essential services like process management and user interaction.

Interrupts

• What Are Interrupts: Interrupts are signals that tell the CPU to stop executing its current instruction and transfer control to an interrupt service routine (ISR).
• Handling Interrupts: The interrupt vector—a table containing pointers to interrupt service routines—helps quickly identify and handle different interrupt types (hardware or software).
• How They Work: The interrupt architecture saves the address of the interrupted instruction. Once the interrupt is handled, the CPU restores its state and resumes the interrupted task.
• Importance: Interrupts allow the system to handle multiple events concurrently, improving the efficiency and responsiveness of the OS.

Storage Structures

RAM and DRAM

• What They Are: RAM (Random Access Memory) is the system’s primary volatile storage, which loses its data when the power is turned off. DRAM (Dynamic RAM) is a type of RAM that requires periodic refresh to maintain its data.
• Purpose: Main memory stores active data and programs that the CPU is currently processing. However, it is not large enough to store all data permanently.
• Importance: Efficient memory management is crucial for performance, as main memory is where active programs and data reside.

Secondary Storage

• What It Is: Secondary storage (e.g., SSDs, HDDs) provides non-volatile storage for data and programs that are not actively in use.
• Types: SSDs are faster but more expensive, while traditional HDDs offer larger storage at a lower price.
• Importance: Secondary storage is vital for storing large amounts of data that exceed the capacity of main memory.

Von Neumann Architecture

• What It Is: This architecture features a single memory system for both data and instructions.
• Components: It uses an instruction register and executes instructions loaded from storage.
• Importance: The Von Neumann architecture is foundational for most modern computing systems.

I/O Structure

I/O Device Controllers

• What They Do: Device controllers manage the interaction between the computer and external devices such as keyboards, disk drives, and printers.
• How They Work: The OS uses device drivers to communicate with the device controllers. These drivers allow the OS to issue commands to devices in a uniform way, simplifying the OS design.
• Interrupt-Driven I/O: For small data transfers, the device controller uses interrupts to notify the OS when an operation is complete. This minimizes CPU involvement and improves system efficiency.
• DMA for Bulk Data Movement: For large data transfers (like disk I/O), Direct Memory Access (DMA) is used. DMA allows data to be transferred directly between memory and device buffers, bypassing the CPU to free it up for other tasks.

Computer System Architecture

Single-Processor Systems

• What They Are: In single-processor systems, there is one primary CPU that executes user instructions.
• Special-Purpose Processors: These systems may also include special-purpose processors, such as disk or keyboard controllers, which handle specific tasks.
• Importance: Single-processor systems are simpler but may have limited performance compared to multi-processor systems.

Multi-Processor Systems

• What They Are: Multi-processor systems use two or more CPUs to handle multiple tasks concurrently.
• Benefits:
  • Increased Throughput: More CPUs mean more tasks can be handled simultaneously.
  • Economy of Scale: The system becomes more cost-efficient as more processors are added.
  • Increased Reliability: Failure of one CPU does not necessarily halt the system; other CPUs can take over the workload.
• Importance: These systems offer better performance, scalability, and fault tolerance.

Reliability of Processors

• Fault Tolerance: Modern systems are designed with fault tolerance in mind, allowing the system to continue functioning even if a component fails.
• Graceful Degradation: Systems with this feature can continue working with reduced functionality, ensuring minimal disruption.
• Importance: High availability is crucial for systems that require continuous operation, such as servers and critical infrastructure.

Asymmetrical vs. Symmetrical Processing

• Asymmetrical Processing: One CPU (the master) controls the system, while the others (slaves) perform specific tasks.
• Symmetrical Processing: All CPUs are equal and share the workload equally.
• Importance: Symmetrical systems are more efficient as they utilize all available processors, whereas asymmetrical systems can be simpler but less efficient.

Memory Management

Memory Access Models

• Uniform Memory Access (UMA): All CPUs have equal access to memory.
• Non-Uniform Memory Access (NUMA): CPUs have different access speeds to different parts of memory.
• Importance: Efficient memory management is crucial for multi-processor systems to avoid bottlenecks and ensure fast data access.

Clustered Systems

What Are Clustered Systems?

• Clustered Systems: These systems consist of multiple CPUs or nodes connected together to form a unified system.
• Types:
  • Asymmetric Clustering: One node is in standby mode, ready to take over if the active node fails.
  • Symmetric Clustering: Multiple nodes are active and monitor each other for failures.
• Importance: Clustering ensures high availability and reliability, making it ideal for applications that require continuous uptime.

Benefits of Clustering:

• High Availability: If one node fails, the others can take over, ensuring minimal disruption.
• Efficiency: Symmetric clustering utilizes all hardware resources, increasing overall system performance.

---

References

Information

• Date: 2025.01.23
• Time: 11:27
• Continued to Computer Organization and Architecture Lecture 11