The stored-program computer

ACompressed recap

🧾

A computer = input, output, memory, ALU, control; CPU = ALU + control + registers. Stored-program: program and data share memory; the PC holds the next instruction's address; instructions run in sequence. Von Neumann (one memory path, the bottleneck) vs Harvard (separate instruction and data memories). Buses: address (one-way), data (two-way), control. Instruction cycle: fetch, advance PC, decode, operand fetch, execute, store.

Performance: T = 1/f; CPU time = IC × CPI × T; CPI of a mix is a weighted average; MIPS = f/(CPI×10⁶); speedup = old/new; Amdahl: 1/((1−F)+F/S), never above 1/(1−F).

BMerged rule card

0.1.1 What a computer is: CPU, memory, I/O, buses; the stored-program idea; fetch–decode–execute

• Five units: input, output, memory, ALU, control. CPU = ALU + control + registers.
• Stored program: instructions and data in one memory; PC = address of the next instruction.
• Von Neumann: one memory path (bottleneck). Harvard: separate instruction and data memories.
• Address bus one-way (CPU → memory); data bus two-way; control bus carries commands and status.
• Instruction cycle: fetch → PC advance → decode → operand fetch → execute → store.

0.1.2 Performance: clock, cycle time, CPI, execution time, MIPS, speedup, Amdahl's law

• T = 1/f. CPU time = IC × CPI × T = IC × CPI / f.
• CPI of a mix = Σ fraction × cycles.
• MIPS = f / (CPI × 10⁶); compare machines by CPU time, not MIPS.
• Speedup = old time / new time. Amdahl: 1 / ((1−F) + F/S); upper limit 1/(1−F).

CInterleaved GATE-style questions

Q1MCQ1 markProgram counter

While an instruction is being executed, the PC normally contains:

• (A) the instruction itself
• (B) the address of the next instruction
• (C) the result of the instruction
• (D) the address of that instruction

Q2NAT1 markCycle time

What is the clock cycle time in ns of a 250 MHz processor?

Numerical answer: ________

Q3NAT2 marksPerformance equation

A program runs in 3 s on a 2 GHz processor with CPI 1.5. How many instructions (in units of 10⁹) does it execute?

Numerical answer: ________

Q4NAT2 marksAmdahl limit

25% of a program's time cannot be parallelised or sped up. Even with infinitely many processors for the rest, what is the maximum overall speedup?

Numerical answer: ________

Q5MCQ1 markData bus

Which bus carries the value read from memory back to the CPU?

• (A) None; memory writes directly into the ALU
• (B) Address bus
• (C) Data bus
• (D) Control bus

Q6NAT2 marksNew clock

A program takes 10 s on a 2 GHz machine. A redesign raises the clock to 3 GHz but increases CPI by 20%. New execution time in seconds?

Numerical answer: ________

Q7MSQ2 marksStatements

Which statements are always true?

• (A) A processor with a higher MIPS rating runs every program faster.
• (B) Halving the cycle time doubles the clock rate.
• (C) Amdahl's law limits speedup to 1/(1−F) for any S.
• (D) CPI depends only on the processor, not on the program.

Q8NAT1 markMemory accesses

A one-word instruction reads two operands from memory and writes its result to a register. How many memory accesses does it make?

Numerical answer: ________

DAnswer Vault