Lesson 1 — Beginner
Computer Organisation & Architecture kya hai
Computer apne andar instruction kaise chalta hai, kaunsa part kya karta hai, aur speed kaise measure hoti hai — ye sab Computer Organisation & Architecture (COA) kehlaata hai. Is lesson mein hum samjhenge ki organisation aur architecture mein kya farak hai, computer ke core components kaun se hain, aur wo formula kya hai jo har exam mein aata hai.
COA kya hai — simple definition
WHAT
Computer Organisation & Architecture computer ke internal design ka wo subject hai jo batata hai machine andar se kaise bani hai aur kaise chalti hai — CPU, memory, I/O, buses aur unka communication. Architecture design ka blueprint hai, organisation us blueprint ka hardware implementation.
WHEN
Jab bhi performance samajhna ho, PC/CPU khareedna ho, koi system design karna ho, ya GATE/UGC-NET/placement ki taiyari ho — COA har jagah base ban jaata hai.
WHERE
Har computer, mobile, server aur embedded device ke andar yahi principles hain. CPU designers, embedded engineers aur har CS student roz isi se deal karte hain.
Ek line mein: COA computer ka anatomy hai — jaise doctor ko body ka structure pata hota hai, waise COA wala ko computer ka andar ka structure. Iska do hissa hai — Architecture (kya-kya hai) aur Organisation (kaise bana hai).
Organisation aur Architecture ka farak
Architecture = wo sab jo PROGRAMMER ko dikhta hai
- Instruction set (ADD, MOV, JMP ke kaun se hain)
- Data types, registers, addressing modes
- Machine language ka format
- Ye badalna matlab naya compiler chahiye
Organisation = us architecture ka HARDWARE implementation
- Control signals kaise bante hain
- Memory technology, bus width, pipelining
- Interconnection kaise hui
- Ye badalne se program wahi chalega (jaise10th gen vs 12th gen i5, same Windows)
Example : dono Intel i5 — architecture same, organisation (cache size,
clock, pipeline depth) alag → speed alag
Simple yaad rakho: Architecture = design, Organisation = building material aur tareeka. Same architecture pe bahut saari organisations ban sakti hain — isliye same "i5" naam ke andar tez aur slow version milte hain.
Computer ke core components
- Control Unit (CU): har instruction ke control signals generate karta hai — computer ka conductor.
- ALU: saari arithmetic (+, −, ×, ÷) aur logic (AND, OR, comparison) calculations yahin hoti hain.
- Registers: CPU ke andar ki sabse fast chhoti memory — PC, IR, ACC yahin hain.
- Main Memory (RAM): programs aur data yahin rehte hain jab tak chal rahe ho.
- I/O devices: duniya se baat — keyboard, monitor, disk sab isi ke through.
- System Bus: teeno (CPU, memory, I/O) ko jodne wali highway — address, data, control lines.
┌──────────────┐
│ CPU │
│ ┌───┐ ┌───┐ │
│ │CU │ │ALU│ │
│ └───┘ └───┘ │
│ Registers │
└──────┬───────┘
System Bus │ (address + data + control)
┌────────────┼────────────┐
│ │ │
┌─────┐ ┌────────┐ ┌─────────┐
│ RAM │ │ Disk │ │I/O (K/B)│
└─────┘ └────────┘ └─────────┘
Performance metrics — speed ka formula
Computer ka performance naapne ka single sabse important formula ye hai — CPU Execution Time = Instruction Count × CPI × Clock Cycle Time. Har numerical isi se shuru hota hai:
CPU Time = IC × CPI × T_clk
IC = kitne instructions chalne hain (instruction count)
CPI = ek instruction average kitne cycles leta hai
T_clk = ek clock cycle ka time (1 / clock frequency)
Example : IC = 5 × 10^8, CPI = 2, clock = 1 GHz (T = 1 ns)
CPU Time = 5×10^8 × 2 × 1ns = 1 second
MIPS = Clock(MHz) / CPI = 1000 / 2 = 500 MIPS
Performance ka simple sense :
IC kam karo → program chhota
CPI kam karo → better instruction design (pipeline)
Clock tez karo → naya hardware
- CPI: Cycles Per Instruction — average kitne clock cycles per instruction. Kam CPI = behtar pipeline/organisation.
- MIPS: Million Instructions Per Second — fast dikhne wala metric, lekin alone misleading (slow instruction zyada ho toh bhi MIPS high aa sakta hai).
- MFLOPS: floating point operations per second — scientific computing mein use.
- Response time vs Throughput: ek task kitna leta hai vs unit time mein kitne tasks.
Exercise
Question: CPU execution time ke teen factors comma se likho (instruction count, cpi, clock cycle time — isi order mein, formula jaisa).
Question: Hardware implementation ke details (control signals, bus width) kis mein aate hain — architecture ya organisation?
Question: CPI ka full form kya hai? (do words, space se)
Common mistakes
- Architecture aur organisation ko same samajhna: Architecture = programmer ko dikhne wala design (ISA), organisation = hardware ka implementation. Exam mein farak poochha jaata hai.
- MIPS ko perfect speed metric samajhna: MIPS instruction count par depend karta hai — do processors ka MIPS same ho sakta hai while performance alag ho.
- Formula mein clock frequency aur time ka confusion: Formula mein cycle TIME aata hai (1/frequency). 1 GHz = 1 ns — seedha multiply nahi, reciprocal lena padta hai.
Cycle ka architecture = usme kya-kya hai (21 gear, disc brake, alloy frame). Organisation = wo parts kaise bane hain (steel ya alloy, kaunsa gear kaise):
Architecture : "21-speed gear, hydraulic brake" (design)
Organisation : "steel frame, 6061 alloy" (implementation)
Same design + alag material = same architecture, alag organisation
Waise hi same i5 + alag cache = same arch, alag org- GATE: execution time formula (IC × CPI × T) par direct numerical aata hai.
- UGC-NET: architecture vs organisation difference classic question.
- TCS NQT: "Cycles Per Instruction" full form + MIPS calculation.
- Trap: "Bus width architecture ka part hai" — galat, wo organisation ka hissa hai.
COA ka matlab samajh aa gaya aur formula bhi haath laga. Agla lesson — Von Neumann Architecture — mein dekhenge computer chalta kya hai wo famous stored-program concept se.