Last Minute Notes Computer Organization

Basic Terminology

• Control Unit - A control unit (CU) handles all processor control signals. It directs all input and output flow, fetches the code for instructions and controlling how data moves around the system.
• Arithmetic and Logic Unit (ALU) - The arithmetic logic unit is that part of the CPU that handles all the calculations the CPU may need, e.g. Addition, Subtraction, Comparisons. It performs Logical Operations, Bit Shifting Operations, and Arithmetic Operation. Figure - Basic CPU structure, illustrating ALU
• Main Memory Unit (Registers) -
  1. Accumulator: Stores the results of calculations made by ALU.
  2. Program Counter (PC): Keeps track of the memory location of the next instructions to be dealt with. The PC then passes this next address to Memory Address Register (MAR).
  3. Memory Address Register (MAR): It stores the memory locations of instructions that need to be fetched from memory or stored into memory.
  4. Memory Data Register (MDR): It stores instructions fetched from memory or any data that is to be transferred to, and stored in, memory.
  5. Current Instruction Register (CIR): It stores the most recently fetched instructions while it is waiting to be coded and executed.
  6. Instruction Buffer Register (IBR): The instruction that is not to be executed immediately is placed in the instruction buffer register IBR.

• Input/Output Devices - Program or data is read into main memory from the input device or secondary storage under the control of CPU input instruction. Output devices are used to output the information from a computer.
  
• Buses - Data is transmitted from one part of a computer to another, connecting all major internal components to the CPU and memory, by the means of Buses. Types:
  1. Data Bus: It carries data among the memory unit, the I/O devices, and the processor.
  2. Address Bus: It carries the address of data (not the actual data) between memory and processor.
  3. Control Bus: It carries control commands from the CPU (and status signals from other devices) in order to control and coordinate all the activities within the computer.

Types of Computer Architecture

1. Von Neumann Architecture

• It uses one memory to store both the program instructions and the data.
• The CPU fetches instructions and data from the same place, one after another.
• This design is simpler and used in most traditional computers.

2. Harvard Architecture

• It uses two separate memories: one for program instructions and another for data.
• The CPU can fetch instructions and data at the same time, making it faster.
• This design is used in modern systems like embedded processors.

Instruction Set and Addressing Modes

Instruction Formats (Zero, One, Two and Three Address Instruction)

A instruction is of various length depending upon the number of addresses it contain. Generally CPU organization are of three types on the basis of number of address fields:

1. Single Accumulator organization
2. General register organization
3. Stack organization

Read more about Instruction Format, Here.

Basic Machine Instructions in COA

Machine instructions are the basic commands given to the processor to perform tasks. They operate directly on the hardware.

Types of Machine Instructions

• Data Transfer Instructions
  • Move data between memory, registers, or I/O devices.
  • Example: LOAD, STORE, MOVE.
• Arithmetic Instructions
  • Perform arithmetic operations like addition, subtraction, multiplication, and division.
  • Example: ADD, SUB, MUL, DIV.
• Logical Instructions
  • Perform logical operations such as AND, OR, NOT, XOR.
  • Example: AND, OR, NOT, XOR.
• Control Transfer Instructions
  • Change the sequence of execution (jump, branch, or call).
  • Example: JUMP, CALL, RET.
• Input/Output Instructions
  • Allow communication between the processor and external devices.
  • Example: IN, OUT.
• Shift and Rotate Instructions
  • Shift or rotate bits in a register.
  • Example: SHL (Shift Left), SHR (Shift Right), ROL (Rotate Left), ROR (Rotate Right).

Components of an Instruction

• Opcode: Specifies the operation to perform (e.g., ADD, SUB).
• Operands: Data to be operated on (e.g., registers, memory locations).

Addressing Modes

The addressing mode specifies a rule for interpreting or modifying the address field of the instruction before the operand is actually executed. An assembly language program instruction consists of two parts :

Effective address or Offset: An offset is determined by adding any combination of three address elements: displacement, base and index.

Read more about Addressing Modes, Here.

RISC vs CISC

Read more about RISC vs CISC, Here.

Instruction Design and Format

CPU Registers

The instruction cycle involves multiple registers in the CPU to fetch, decode, execute and store results.

• Program Counter (PC)
  • Holds the address of the next instruction to be executed.
  • Updates after each instruction fetch.
• Instruction Register (IR)
  • Stores the currently fetched instruction.
  • Used by the control unit for decoding.
• Memory Address Register (MAR)
  • Holds the memory address of the data or instruction to be fetched or stored.
• Memory Data Register (MDR) (or Memory Buffer Register, MBR)
  • Temporarily holds the data being transferred to/from memory.
• Accumulator (AC)
  • Stores intermediate arithmetic and logic results during execution.
• General Purpose Registers (GPR)
  • Temporary storage for operands, results, or data during execution.
• Temporary Register (TR)
  • Stores intermediate data during complex operations or instruction execution.
• Status Register / Flag Register
  • Stores condition flags (e.g., zero, carry, overflow) to indicate the result of operations.
• Stack Pointer (SP)
  • Points to the top of the stack in memory, used during function calls or interrupts.
    

Flag Registers

Status Flags

• Zero Flag (Z): When an arithmetic operation results in zero, the flip-flop called the Zero flag - which is set to one.
• Carry flag (CY): After an addition of two numbers, if the sum in the accumulator is larger than eight bits, then the flip-flop uses to indicate a carry called the Carry flag, which is set to one.
• Parity (P): If the result has an even number of 1s, the flag is set to 1; for an odd number of 1s the flag is reset.
• Auxiliary Carry (AC): In an arithmetic operation, when a carry is generated from lower nibble and passed on to higher nibble then this register is set to 1.
• Sign flag(S): It is a single bit in a system status (flag) register used to indicate whether the result of the last mathematical operation resulted in a value in which the most significant bit was set.

Instruction Cycle

1. Fetch: The CPU retrieves the next instruction from memory using the Program Counter (PC).

2. Indirect: If the instruction uses an indirect addressing mode, the effective memory address of the operand is resolved. Example: For LOAD R1, (100), the CPU fetches the address stored at memory location 100.

3. Execute: The CPU performs the operation specified by the instruction (e.g., arithmetic, logical, control).

4. Interrupt: If an interrupt request occurs (e.g., hardware interrupt or software exception), the CPU temporarily halts the current execution to service the interrupt. After servicing, the CPU resumes the instruction cycle.

Standard Instruction Cycle

This includes the basic steps for executing instructions:

• Fetch: Retrieve the instruction from memory.
• Decode: Identify the operation and operands.
• Execute: Perform the operation.
• Store (Write Back): Save the result (if any).

Read more about Instruction Cycle, Here.

Control Unit

The Control Unit (CU) is a core component of the CPU that directs its operation by generating control signals. It manages the execution of instructions by coordinating with the ALU, registers, and memory.

Types of Control Units

Hardwired Control Unit -

• Fixed logic circuits that correspond directly to the Boolean expressions are used to generate the control signals.
• Hardwired control is faster than micro-programmed control.
• A controller that uses this approach can operate at high speed.
• RISC architecture is based on hardwired control unit.

Micro-programmed Control Unit -

• The control signals associated with operations are stored in special memory units inaccessible by the programmer as Control Words.
• Control signals are generated by a program are similar to machine language programs.
• Micro-programmed control unit is slower in speed because of the time it takes to fetch microinstructions from the control memory.
• There are two type Micro-programmed control Unit:
  • Horizontal Micro-programmed control Unit- The control signals are represented in the decoded binary format that is 1 bit/CS.
  • Vertical Micro-programmed control Unit - The control signals re represented in the encoded binary format. For N control signals- Log_n(N) bits are required.

Read more about Hardwired CU vs Micro-programmed CU, Here.

Microprogram: Program stored in memory that generates all control signals required to execute the instruction set correctly, it consists micro-instructions.

Micro-instruction: Contains a sequencing word and a control word. The control word is all control information required for one clock cycle.

Micro-operations: Micro-operations are the atomic operations which executes a particular micro-instruction. Example of micro-operation during the fetch cycle:

t1: MAR ←(PC)
t2: MBR ←Memory
    PC ←(PC) + I
t3: IR ←(MBR)

Memory Organization

• Memories are made up of registers.
• Each register in the memory is one storage location.
• The storage location is also called a memory location.
• Memory locations are identified using Address.
• The total number of bit a memory can store is its capacity.

Memory Hierarchy

Simultaneous access memory organization: If H1 and H2 are the Hit Ratios and T1 and T2 are the access time of L1 and L2 memory levels respectively then the

Average Memory Access Time can be calculated as:

T=(H1*T1)+((1-H1)*H2*T2

Hierarchical Access Memory Organization: If H1 and H2 are the Hit Ratios and T1 and T2 are the access time of L1 and L2 memory levels respectively then

Average Memory Access Time can be calculated as:

T=(H1*T1)+((1-H1)*H2*(T1+T2)

Read more about Simultaneous and Hierarchical Access Memory Organization, Here.

Cache Memory

Cache Memory is a special very high-speed memory. It is used to speed up and synchronizing with high-speed CPU. Levels of memory: Level 1 or Register, Level 2 or Cache memory, Level 3 or Main Memory, Level 4 or Secondary Memory.

Hit ratio = hit / (hit + miss) = no. of hits/total accesses

Locality of reference - Since size of cache memory is less as compared to main memory. So to check which part of main memory should be given priority and loaded in the cache is decided based on the locality of reference.

Types of Locality of reference

• Spatial Locality of reference: Spatial locality means instruction or data near to the current memory location that is being fetched, may be needed soon in the near future.

• Temporal Locality of reference: Temporal locality means current data or instruction that is being fetched may be needed soon. So we should store that data or instruction in the cache memory to avoid searching again in main memory for the same data.

• Cache Mapping: There are three different types of mapping used for the purpose of cache memory which is as follows: Direct mapping, Associative mapping and Set-Associative mapping.

Direct Mapping - Maps each block of main memory into only one possible cache line. If a line is previously taken up by a memory block and a new block needs to be loaded, the old block is trashed. An address space is split into two parts index field and a tag field. The cache is used to store the tag field whereas the rest is stored in the main memory.

Cache Line Number = Main Memory block Number % Number of Blocks in Cache

Associative Mapping - A block of main memory can map to any line of the cache that is freely available at that moment. The word offset bits are used to identify which word in the block is needed, all of the remaining bits become Tag.

Set-Associative Mapping - Cache lines are grouped into sets where each set contains k number of lines and a particular block of main memory can map to only one particular set of the cache. However, within that set, the memory block can map to any freely available cache line.

Cache Set Number = Main Memory block number % Number of sets in cache

Note: Translation Lookaside Buffer (i.e. TLB) is required only if Virtual Memory is used by a processor. In short, TLB speeds up the translation of virtual address to a physical address by storing page-table in faster memory. In fact, TLB also sits between the CPU and Main memory.

Read more about Cache Mapping Techniques, Here.

Multilevel Cache

Multilevel Caching is used in modern processors to improve memory access speed by introducing multiple levels of cache memory.

Types of Cache Levels

• L1 Cache (Level 1):
  • Smallest, fastest, and closest to the CPU.
  • Usually divided into Instruction Cache and Data Cache.
• L2 Cache (Level 2):
  • Larger and slower than L1 but still faster than main memory.
  • Acts as a bridge between L1 and L3/main memory.
• L3 Cache (Level 3):
  • Shared across multiple cores.
  • Larger and slower than L2 but faster than main memory.

Performance Metrics

1. Hit Ratio:Percentage of memory accesses satisfied by the cache. \text{Hit Ratio} = \frac{\text{Cache Hits}}{\text{Total Accesses}}
2. Miss Ratio: Percentage of memory accesses that result in a miss. \text{Miss Ratio} = 1 - \text{Hit Ratio} Effective Memory Access Time (EMAT)
   For 2-level cache:
   \text{EMAT} = H_1 \times T_1 + (1 - H_1) \times [H_2 \times T_2 + (1 - H_2) \times T_M]
   • H1,H2H_1, H_2: Hit ratios for L1 and L2 caches.
   • T1,T2T_1, T_2: Access times for L1 and L2 caches.
   • TMT_M: Access time for main memory.

Cache Replacement Policies Table

Cache Updation Policy

Write Through: In this technique, all write operations are made to main memory as well as to the cache, ensuring that main memory is always valid.

For hierarchical access: T_{read} = H \times T_{cache} + (1-H) \times (T_{cache} + T_{memory\_block}) \newline= T_{cache} + (1-H) \times T_{memory\_block}

For simultaneous access : T_{read} = H \times T_{cache} + (1-H) \times (T_{memory\_block}) \newlineT_{write} = T_{memory\_word}

Write Back: In write-back updates are made only in the cache. When an update occurs, a dirty bit, or use bit, associated with the line is set. Then, when a block is replaced, it is written back to main memory if and only if the dirty bit is set.

For hierarchical access: T_{read} = T_{write} = H \times T_{cache} + (1-H) \times (T_{cache} + T_{memory\_block} + T_{write\_back}) \\= T_{cache} + (1-H) \times (T_{memory\_block} + T_{write\_back}), \\ \text{ where } T_{write\_back} = x \times T_{memory\_block}, \text{ where } x \text{ is the fraction of dirty blocks}

For simultaneous access : T_{read} = T_{write} = H \times T_{cache} + (1-H) \times ( T_{memory\_block} + T_{write\_back}), \\ \text{ where } T_{write\_back} = x \times T_{memory\_block}, \text{ where } x \text{ is the fraction of dirty blocks}

Read more about Cache Memory, Here.

Cache Miss

Read about Types of Cache Miss, Here.

I/O Interface

• An I/O (Input/Output) Interface connects the CPU and memory with external devices like keyboards, monitors, printers, etc.
• It acts as a bridge between the CPU and I/O devices to ensure smooth data transfer.

Data transfer between the main memory and I/o device may be handled in a variety of modes like :

Programmed I/O: In Programmed I/O, the CPU controls data transfer between the I/O device and memory without allowing direct access for the device. The I/O device sends one byte at a time, placing the data on the I/O bus and enabling the data valid line. The interface stores the byte in its data register, activates the data accepted line, and sets a flag bit to notify the CPU. The I/O device waits for the data accepted line to reset before sending the next byte. This process is managed step-by-step by the CPU, making it slower but synchronized.

Interrupt driven I/O: In interrupt driven I/O, the processor issues an I/O command, continues to execute other instructions, and is interrupted by the I/O module when the I/O module completes its work.

Read more about Interrupt, Here.

Interrupt Handling Techniques

• Daisy Chaining in Interrupts

Daisy chaining is a method of handling multiple interrupts in a system by connecting the devices in a serial or chain-like manner. When an interrupt request is generated, the priority is determined by the position of the device in the chain. The device closer to the CPU has higher priority. The interrupt signal travels through the chain, and each device checks if it is the source of the interrupt. If not, it passes the signal to the next device in the chain. This approach is simple to implement but suffers from longer delays for devices farther down the chain and is unsuitable for systems requiring precise or equal priority handling.

• Parallel Priority Interrupt

Parallel priority interrupts use a priority encoder to handle multiple interrupt requests simultaneously. All devices send their interrupt requests in parallel to the encoder, which determines the highest-priority interrupt and sends it to the CPU. This method is faster and more efficient than daisy chaining because it does not rely on signal propagation through a chain. Each device is assigned a priority, and the encoder ensures that the device with the highest priority gets serviced first. Parallel priority interrupts are commonly used in systems where speed and fair priority handling are essential.

Direct Memory Access(DMA): In Direct Memory Access (DMA), the I/O module and main memory exchange data directly without processor involvement.

Modes of DMA Transfer

1. Burst Mode (Block Transfer Mode)

In burst mode, the DMA controller takes full control of the system bus and transfers an entire block of data in one go before releasing the bus back to the CPU. This method is fast but can cause the CPU to be idle during the transfer, as it doesn't get access to the bus until the transfer is complete.

2. Cycle Stealing Mode

In cycle stealing mode, the DMA controller takes control of the bus for one data transfer (one word or one byte) at a time and then releases it back to the CPU. This allows the CPU and DMA to share the bus alternately, improving overall system efficiency while slightly slowing the DMA transfer.

Read more about Modes of DMA Transfer, Here.

DMA Controller

The DMA (Direct Memory Access) Controller is a hardware component that manages data transfer between memory and I/O devices without constant CPU involvement. It communicates with the CPU, memory, and I/O devices through control and data lines.

The CPU interacts with the DMA controller by selecting its registers via the address bus while enabling the DS (Data Select) and RS (Register Select) inputs. When the CPU grants the bus to the DMA (indicated by BG = 1, Bus Grant), the DMA takes control of the buses. The DMA then directly communicates with memory by placing the memory address on the address bus and activating the RD (Read) or WR (Write) control signals to perform data transfer.

The DMA controller communicates with external I/O devices using request and acknowledge lines:

• The I/O device sends a request signal when it needs to transfer data.
• The DMA acknowledges this request, initiates the data transfer, and ensures synchronization.

This process enables efficient and high-speed data transfer while freeing the CPU to perform other tasks.

Read more about I/O Interface, Here.

Pipelining

• Pipelining is a process of arrangement of hardware elements of the CPU such that its overall performance is increased.
• Simultaneous execution of more than one instruction takes place in a pipelined processor.
• RISC processor has 5 stage instruction pipeline to execute all the instructions in the RISC instruction set. Following are the 5 stages of RISC pipeline with their respective operations:
  • Stage 1 (Instruction Fetch) In this stage the CPU reads instructions from the address in the memory whose value is present in the program counter.
  • Stage 2 (Instruction Decode) In this stage, instruction is decoded and the register file is accessed to get the values from the registers used in the instruction.
  • Stage 3 (Instruction Execute) In this stage, ALU operations are performed.
  • Stage 4 (Memory Access) In this stage, memory operands are read and written from/to the memory that is present in the instruction.
  • Stage 5 (Write Back) In this stage, computed/fetched value is written back to the register present in the instructions.

Performance of a pipelined processor

Consider a 'k' segment/stages pipeline with clock cycle time as 'Tp'. Let there be 'n' tasks to be completed in the pipelined processor. So, time taken to execute 'n' instructions in a pipelined processor:

ET_pipeline = k + n – 1 cycles
= (k + n – 1) Tp

In the same case, for a non-pipelined processor, execution time of 'n' instructions will be:

ET_non-pipeline = n * k * Tp

So, speedup (S) of the pipelined processor over non-pipelined processor, when 'n' tasks are executed on the same processor is:

S = Performance of pipelined processor / Performance of Non-pipelined processor

As the performance of a processor is inversely proportional to the execution time, we have:

S = ET_non-pipeline / ET_pipeline
=> S = [n * k * Tp] / [(k + n – 1) * Tp]
S = [n * k] / [k + n – 1]

When the number of tasks 'n' are significantly larger than k, that is, n >> k

S = n * k / n
S = k

where 'k' are the number of stages in the pipeline. Also,

Efficiency = Given speed up / Max speed up = S / S_max

We know that, Smax = k So,

Efficiency = S / k

Throughput = Number of instructions / Total time to complete the instructions So,

Throughput = n / (k + n – 1) * Tp

Note: The cycles per instruction (CPI) value of an ideal pipelined processor is 1

Performance of pipeline with stalls

Speed Up (S) = CPI_non-pipeline/ (1 + Number of stalls per instruction)

Read more about Pipelining, Here.

Dependencies and Data Hazard

There are mainly three types of dependencies possible in a pipelined processor. These are :

Structural dependency:

• This dependency arises due to the resource conflict in the pipeline. A resource conflict is a situation when more than one instruction tries to access the same resource in the same cycle. A resource can be a register, memory, or ALU.
• To minimize structural dependency stalls in the pipeline, we use a hardware mechanism called Renaming.

Control Dependency:

• This type of dependency occurs during the transfer of control instructions such as BRANCH, CALL, JMP, etc. On many instruction architectures, the processor will not know the target address of these instructions when it needs to insert the new instruction into the pipeline. Due to this, unwanted instructions are fed to the pipeline.
• Branch Prediction is the method through which stalls due to control dependency can be eliminated. In this at 1st stage prediction is done about which branch will be taken.

Data Dependency :

• Data dependency occurs when one instruction depends on the result of another instruction. It can cause data hazards in pipelined processors.

Types of Hazards in Pipelined Processors

Hazards are situations that cause the pipeline to stall or delay instruction execution. There are three main types of hazards:

1. Structural Hazards

• Occur when hardware resources are insufficient to handle the current instruction stream.
• Example: If only one memory unit exists, and both instruction fetch and data access need it simultaneously.

Solution:

• Add more resources (e.g., separate instruction and data memory – Harvard architecture).
• Use scheduling to avoid conflicts.

2. Data Hazards

• Arise when an instruction depends on data from a previous instruction that has not yet completed.

Types of Data Hazards:

• RAW (Read After Write): True dependency.
• WAR (Write After Read): Anti-dependency.
• WAW (Write After Write): Output dependency.

Solution:

• Data forwarding/bypassing.
• Insert pipeline stalls (NOPs).
• Instruction scheduling.

3. Control Hazards

• Occur due to branch or jump instructions, where the next instruction to execute is uncertain until the branch is resolved.

Solution:

• Branch prediction techniques.
• Delayed branching (use NOPs).
• Dynamic scheduling.

Read more about Dependencies and Hazards, Here.

IEEE Standard 754 Floating Point Numbers

There are several ways to represent floating point number but IEEE 754 is the most efficient in most cases. IEEE 754 has 3 basic components:

The Sign of Mantissa - This is as simple as the name. 0 represents a positive number while 1 represents a negative number.

The Biased exponent - The exponent field needs to represent both positive and negative exponents. A bias is added to the actual exponent in order to get the stored exponent.

The Normalised Mantisa - The mantissa is part of a number in scientific notation or a floating-point number, consisting of its significant digits. Here we have only 2 digits, i.e. O and 1. So a normalised mantissa is one with only one 1 to the left of the decimal.

The IEEE 754 Standard is used to represent floating-point numbers in binary. It has two formats:

1. Single Precision (32-bit)
2. Double Precision (64-bit)

E=0,M=0: Zero

Read more about IEEE Floating Point Notation, Here.