Basic Computer System Architecture - DA 107

Course Information

• Course title: Bachelor of Science (Honours) in Data Science and Artificial Intelligence
• Course code: DA 107
• Module title: Basic Computer System Architecture

Introduction to the Module (Page 3 & Page 4)

• The module introduces the fundamental components and operations of a computer system.
• The outline covers various concepts, including first-generation computers, computer architecture, basic computing concepts, registers, RAM, and more

First-Generation Computers & Architecture (Page 8)

• The Electronic Numerical Integrator and Computer (ENIAC) was designed by John Mauchley and J. Presper Eckert.
• ENIAC had 18,000 vacuum tubes and 1,500 relays.
• It weighed 30 tons and consumed 140 kW of power.
• ENIAC had 20 registers, capable of holding 10-digit decimal numbers.
• Programming involved setting up 6000 multi-position switches and connecting sockets.
• ENIAC was built around 1946.

First-Generation Computers & Architecture (Page 9)

• The Electronic Discrete Variable Automatic Computer (EDVAC) is the successor to ENIAC.
• John von Neumann designed an improved version of EDVAC.
• Von Neumann realised programming computers with numerous switches and cables was inefficient.
• His design allowed programs and data to be stored in computer memory, making the system more flexible.
• He proposed replacing the decimal arithmetic with parallel binary arithmetic.

First-Generation Computers & Architecture (Page 10)

• The basic design of EDVAC, now called the von Neumann machine, is still used in many computers.
• It was used in the first stored-program computer EDSAC and is still used a half a century later.

First-Generation Computers & Architecture (Page 11)

• Computer memory consisted of 4096 words with each word holding 40 bits.
• Each word held two 20-bit instructions or a 40-bit signed integer.
• Instructions involved 8 bits for instruction type and 12 bits to specify one of the 4096 memory words.
• The ALU (Arithmetic Logic Unit) and the control unit form the "brain" of the computer.
• The ALU and the control unit are combined on a single chip called the CPU.

Basic Computing Concepts (Pages 13-16)

• The microprocessor, or CPU, is the central processing unit.
• A computer processes input instructions and data.
• Instructions include addition, subtraction, multiplication, and division.
• The result of these operations form an output stream.
• The fundamental operations include reading, writing, and modification of data stored in computer memory.
• These functions rely on computer components like storage, ALU, and a bus system.

Basic Computing Concepts (Page 17)

• Computers need a storage structure to hold numbers that can be accessed for reading and writing.

Basic Computing Concepts (Page 18)

• The ALU performs arithmetic operations upon the input data.
• Numbers are transferred to the ALU from memory through the data bus.
• The result of the operations is returned to memory via the ALU's output port.

Basic Computing Concepts (Page 19)

• The bus serves as a communication network to move data between ALU and memory.
• Data that moves between the ALU and memory, and instructions also use the bus.
• The instruction bus is specifically for transferring instructions.

Basic Computing Concepts (Page 20)

• The ALU processes data according to specific steps and instructions.
• Steps for performing addition include obtaining input numbers, performing the addition, and storing the result into memory.
• Modern CPUs perform these types of operations billions of times a second.

The Register File (Pages 21-25)

• Computers use specialized storage locations, registers, close to the ALU.
• Registers are faster than memory and provide immediate access to data required in computation.
• The x86 architecture, for instance, utilizes a register file with 8 registers.
• Operations in the ALU, such as addition, use registers for input operands and results.

The Register File (Page 26)

• Instructions and data exist in specific registers and memory is used to store these variables and instructions.
• Operations in a simple computer can involve 4 registers A, В, С, D, to store numbers and carry out arithmetic operations.
• These registers will store the result of operations and overwrite previous values.

The Register File (Page 27)

• A simple description of the process of performing addition of two numbers stored in registers A and В into register C.

RAM (Pages 28-30)

• RAM, or Random Access Memory, is a large storage space.
• It handles data that is not stored in registers.
• A portion of data is moved to registers for easier processing by the ALU.
• Main memory handles data which is not immediately required by the ALU, it is transferred to the registers temporarily for calculations.

RAM (Page 31)

• How a computer performs arithmetic operations involving main memory and registers (using the CPU and ALU).
• Arithmetic operations involve reading two operands from main memory and loading them into registers.
• The ALU adds values from two source registers and store the result in the destination register.
• Then the destination register is stored in main memory.

RAM (Page 32)

• The user needs to control data flow to and from main memory and CPU registers to manage the execution of operations.
• Instructions for memory operations also need to be given to the CPU and not only the ALU.

A Closer Look At The Code Stream (Pages 35&36)

• Computer programs are sequences of instructions, which are often in binary formats.
• These command all the computer’s actions, not just the ALU actions alone.
• Instructions encompass more than simple arithmetic, like operations on data and memory.
• Instructions can be categorized into arithmetic instructions (like addition, subtraction) and memory access instructions (like loading from memory and storing to memory).

A Closer Look At The Code Stream (Page 37)

• DLW-1 is a hypothetical computer for illustration.
• It has 4 registers: A, B, C, D.
• It has 256 memory cells, numbered from 0 to 255, to store variables.

DLW-1's Arithmetic Instruction Format (Page 39)

• Instruction format in DLW-1 involves an opcode, that specifies the operation, and source registers and a destination register. This format is often written in hexadecimal.

DLW-1's Memory Instruction Format (Page 39)

• Memory access instructions, like load and store, specify the memory location to read data from, or the location to write data to.

Example DLW-1 Program (Page 41)

• Example code and its binary machine language equivalents, and a program for adding two integers stored in different memory locations.

Register Vs Immediate (Pages 42-44)

• Programmers need to know memory locations to access and modify data.
• It is often impractical to access memory locations directly due to billions of possible addresses.
• Modern computers use the concept of addressing data and instructions stored in registers that can be used as temporary storage to access memory locations quickly.

Register Vs Immediate (Page 45)

• Immediate values are used instead of memory locations directly.

Register Vs Immediate (Pages 45-47)

• Number 12 and 14 are stored in register D, and instructions to access and operate on this data and store the output in a memory location, using registers and immediate values.

Relative Addressing (Page 49)

• Data segments are blocks of memory cells.
• Relative addressing is a method to find memory cells within a segment using an offset from a base address.
• Memory cell addresses can be obtained using the formula: address = base address + offset.

DLW-1 Program With Use Of Relative Addressing (Page 51)

• Instructions like load and store use relative addressing and calculating memory addresses
• Instructions specify the base address and offset and calculate the memory address to access data in memory

Mechanics of Program Execution (Pages 53-55)

• Instructions must be represented in binary notation (opcodes).
• A binary code, opcode, is a unique group of binary numbers that maps to a computer instruction.
• 3-bit opcodes represent different operations for instructions.
• The complete numerical code for every instruction uses many bits to represent commands, opcodes, location and data in the machine language

Binary Encoding of Arithmetic Instructions (Pages 57-62)

• Arithmetic instructions are expressed in a specific binary format: mode,opcode, source1,source2, destination.
• A 0 for mode gives the register type format.
• An instruction can specify the operation to be performed, which registers hold the input data and where the result should be placed.
• The binary codes and formats change depending upon whether they are register-type or immediate-type operations

Binary Encoding of Arithmetic Instructions With Immediate Value (Pages 63-67)

• Arithmetic operations using immediate values, numbers directly embedded into an instruction.
• Instructions specify a numerical operand and operation, and destination registers.

Binary Encoding of Memory Access Instructions (Pages 68-73)

• Memory access instructions involve loading and storing data from memory to registers, and vice versa.
• They use both register-type instructions and immediate values, specifying registers or addresses in memory.

The Store Instruction (Page 75)

• The binary format for a store instruction is almost identical to that of a load instruction, but the destination field specifies the register where the data will be stored.

Example Programs (Page 79)

• Assembly language instructions and corresponding machine language codes are shown for common operations like loading data, performing addition, and storing results.