Presentation is loading. Please wait.

Presentation is loading. Please wait.

Basic Computer The following discussions are based on a fictitious computer called “Basic Computer” by the author of the textbook It’s a much better way.

Similar presentations


Presentation on theme: "Basic Computer The following discussions are based on a fictitious computer called “Basic Computer” by the author of the textbook It’s a much better way."— Presentation transcript:

1 Basic Computer The following discussions are based on a fictitious computer called “Basic Computer” by the author of the textbook It’s a much better way to learn computer architecture concepts than trying to understand the Intel Pentium architecture

2 Assembly Language Every computer architecture (or family of architectures) has its own unique assembly language Unlike Java, you should not learn assembly language syntax, data types, etc. You should learn to program/think at the assembly language level –It’s a way of thinking that requires intimate knowledge of the underlying hardware architecture

3 Assembly Language Instructions Each instruction has two basic parts –Operation code (opcode) What the instruction wants the processor to do –Operand(s) (registers, memory addresses) Data location that the instruction wants the processor to manipulated Some operands will be explicit while others will be implicit (implied by the opcode)

4 Assembly Language Instructions n-bit instruction format Example – 16 bit instruction opcodeoperand/address 0m+1mn-1 opcodeoperand/address 0121115 2 (n-1)-(m+1) opcodes 2 (m+1) addresses 2 4 = 16 opcodes 2 12 =4096 addresses

5 Assembly Language Instructions Instructions within the same Assembly language may be of differing lengths –i.e. not all instructions utilize the same number of bits as we saw with the Pentium

6 Internal Operation To execute an assembly language instruction the processor goes through 4 steps –Fetch an instruction from memory –Decode the instruction –Read the operands from memory/registers –Execute the instruction This is often referred to as the Fetch-Execute cycle or the Instruction cycle To execute a program the processor repeats this cycle until a halt instruction is reached

7 Internal Operation All this is under the control of the Control Unit This is the component that decodes the instruction and sends out microoperations to the rest of the hardware –The control unit can be hardwired Made up entirely of sequential circuits designed to do precisely the fetch-execute steps – fixed instruction set –The control unit can be microprogrammed A small programmable processor within the processor – programmable instruction set More on this later

8 Addressing Modes In designing a computer architecture the designer must specify the instruction set –Opcode/operand pairs In specifying operands there are a number of alternatives –Immediate instructions –Direct address operands –Indirect address operands

9 Immediate Instruction The 2 nd part of the instruction is the operand (rather than the address of the operand) An example might be an instruction that adds a constant to a register add 3 –The “3” is the value we want to add, not an address in memory

10 Direct Address Instruction The 2 nd part of the instruction is the memory address of operand An example might be an instruction that adds a value in memory to a register add 0x30213 –The “0x30213” is the memory address of the value that we want to add

11 Indirect Address Instruction The 2 nd part of the instruction is the memory address of the location that holds the memory address of the operand An example might be an instruction that adds a value in memory to a register add 0x30213 –The “0x30213” is a memory address that holds the memory address of the value that we want to add

12 Addressing Modes Iopcodeaddress 0addc3 0add0x331add0x33 0x42 0x33 0x42 0x33 0x88 0x42 Operand Immediate DirectIndirect Mode bit

13 Addressing Modes The term effective address refers to the actual address of the operand –For the previous example Immediate address mode –Effective address is the instruction itself Direct address mode –Effective address is the memory location 0x33 Indirect addressing mode –Effective address is the memory location 0x42

14 Addressing Modes Something in the instruction word will specify which addressing mode is applicable –The operand itself (for immediate instructions) –A designated bit (for direct vs. indirect address instructions)

15 Addressing Modes Indirect addressing is a convenient way to implement arrays (which are nothing more than pointers to blocks of contiguous memory) Some architectures define additional modes such as “read location then increment” –These are all derivations of the three defined here

16 Registers In designing a computer architecture the designer must specify the register set There are essentially two categories –Special purpose registers –General purpose registers

17 Special Purpose Registers Program Counter (PC) –Holds the memory address of the next instruction of our program Memory Address Register (AR) –Holds the address of a location in memory that we want to access (read/write) The size of (number of bits in) these two registers is determined by the number of memory addresses in our architecture

18 Special Purpose Registers Instruction Register (IR) –Holds the instruction (opcode/operand) we are about to execute Data Register (DR) –Holds the operand read from memory to be sent to the ALU Accumulator (AC) –Holds an input to the ALU and the output from the ALU

19 Special Purpose Registers Input Register (INPR) –Holds data received from a specified external device Output Register (OUTR) –Holds data to be sent to a specified external device

20 General Purpose Registers Temporary Register (TR) –For general usage either by our program or the architecture

21 Registers These registers (shown previously) are specified for the fictitious architecture given in the textbook All architectures will have these in some form Most architectures will have more than just these –More general purpose registers –Stack pointers –Interrupts –Program status bits –Multiple I/O ports –Timers –etc. To effectively program the architecture (in assembly language) you need to be aware of all the available registers and their usage High level language compilers possess this knowledge

22 Bus In designing a computer architecture the designer must specify the bus layout –The size of the bus (in bits) –What is connected to the bus –Access control to the bus Recall that a bus is an efficient alternative to lots of wires when it comes to transferring data between registers, control units, and memory locations

23 Bus Architecture Memory unit 4096x16 AR PC DR AC INPR IR TR OUTR ALU E 16-bit Bus address clock S2S2 S1S1 S0S0 111 001 010 011 100 101 110 Access Select

24 Bus Architecture The three access select lines determine which register is allowed to write to the bus at a given time (recall that only one write at a time is allowed) Registers have load input signals (LD) that tell them to read from the bus If registers are smaller than the bus (less bits) than unused bits are set to 0 Some registers have additional input signals –Increment (INR) and Clear (CLR) –See figure 5-4, page 130 of the textbook

25 Bus Architecture Memory has read/write input signals that tell it when to take data from the bus and send data to the bus Memory addresses (for both read and write operations) are always specified via the Address Register (AR) –An alternative (used in many architectures) is a two bus system One address bus One data bus

26 Bus Architecture Results of all ALU (arithmetic, logic, and shift operations) are always sent to the Accumulator (AC) register –The ALU is the only way to set values into the accumulator except for the clear (CLR) and increment (INR) control lines Inputs to the ALU come from –The Accumulator (AC) –The Data Register (DR) –The Input Register (INPR) The E output from the ALU is the carry-out (Extended AC) bit –Many architectures pack this into a register with other status bits such as overflow

27 Bus Architecture Some pairs of microoperations can be performed in a single clock cycle –The key is to make sure they don’t both try to put data on the bus –Consider the RTL statement DR ← AC, AC ← DR –This is allowed since the DR ← AC microoperation uses the bus while the AC ← DR microoperation does not

28 Instructions We said previously that there are two parts to an instruction –Opcode –Operand Realistically the two parts should be called –Opcode –Everything else

29 Instructions Three basic types –Those that reference memory operands –Those that reference register operands –Those that reference I/O devices Again, this is only for the fictitious architecture in the textbook but you will find similar categorizations in real architectures

30 Memory Instructions Iopcodeaddress 011121415 I = 0 means direct memory address I = 1 means indirect memory address There are 14 instructions in this class –7 direct memory address forms –7 indirect memory address forms

31 Memory Instructions Hex Code SymbolI = 0I = 1Description AND0xxx8xxxMem AND AC ADD1xxx9xxxMem + AC LDA2xxxAxxxLoad AC from Mem STA3xxxBxxxStore AC to Mem BUN4xxxCxxxUnconditional Branch BSA5xxxDxxxBranch to Subroutine ISZ6xxxExxxIncrement and Skip if Zero

32 Register Instructions There are 12 instructions in this class –They can use the “operand field” to specify the register and type of operation since no memory address is required 0111 Register operation 011121415

33 Register Instructions SymbolHex CodeDescription CLA7800Clear AC CLE7400Clear E bit CMA7200Complement AC CME7100Complement E bit CIR7080Circulate right AC and E CIL7040Circulate left AC and E INC7020Increment AC

34 Register Instructions (cont.) SymbolHex CodeDescription SPA7010Skip next instruction if AC is positive SNA7008Skip next instruction if AC is negative SZA7004Skip next instruction if AC is 0 SZE7002Skip next instruction if E is 0 HLT7001Halt

35 I/O Instructions There are 6 instructions in this class –They can use the “operand field” to specify the exact operation since no memory address is required 1111 I/O operation 011121415

36 I/O Instructions SymbolHex CodeDescription INPF800Input character to AC OUTF400Output character from AC SKIF200Skip on input flag SKOF100Skip on output flag IONF080Interrupt on IOFF040Interrupt off

37 Instruction Decoding The control unit evaluates bits 15 – 12 to determine the instruction format At first glance it appears that there can be only 8 unique instructions since the opcode resides in 4 bits But, additional instructions are created through the use of the I bit an unused bits in the operand field

38 Instruction Set Design To be useful, an architecture’s instruction set must contain enough instructions to allow all possible computations Four categories are necessary –Arithmetic, logical, shift operations –Moving data to/from memory from/to registers –Control such as branch and conditional checks –Input/output

39 Instruction Set Design The set in the book is complete in that all the possible operations on binary numbers can be performed through combinations of instructions But, the set is very inefficient in that highly used operations require multiple instructions This is why the Pentium instruction set is so large and complicated – it makes for efficient programs


Download ppt "Basic Computer The following discussions are based on a fictitious computer called “Basic Computer” by the author of the textbook It’s a much better way."

Similar presentations


Ads by Google