Stretch tasks Hardware and the Processor

Round 1: Quiz Blitz

1. Instructions and data are held in the same memory It is why one laptop can be a browser and a game with no physical change.
2. The address of the next instruction It is incremented during every fetch.
3. MAR MAR holds an address. MDR holds data.
4. Accumulator The ALU puts its result in the accumulator.
5. Control unit It sends the control signals that coordinate everything else.
6. True Straight after its address is copied into the MAR.
7. 3 billion Giga means billion.
8. Some tasks are sequential, and much software cannot split its work A step that needs the last step's result cannot run in parallel.
9. False It reduces the time spent waiting for RAM.
10. Instructions and data compete for one route to memory The processor spends part of its time waiting.
11. A washing machine controller; A traffic light controller An embedded system does one dedicated task inside a larger device.
12. The memory data register It comes back along the data bus into the MDR.

Round 2: Quick Fire

1. • The MAR holds the address being accessed
   • The MDR holds the data or instruction being moved
   Model answer

   The MAR holds the address of the memory location currently being accessed. The MDR holds the data or instruction being moved to or from that location.

2. • More pulses per second means more instructions completed each second
   • Heat and power rise faster than speed
   Model answer

   Each clock pulse completes one stage of the cycle, so more pulses per second means more instructions completed each second. But heat and power consumption rise faster than speed, which is why clock speeds stopped climbing.

3. • It holds frequently used data and instructions close to the processor
   • They are reached faster than RAM, so the CPU waits less
   Model answer

   Cache holds recently and frequently used data and instructions close to the processor. They can be reached far faster than from RAM, so the CPU spends less time waiting and completes more instructions each second.

4. • It does one dedicated task that never changes
   • So it can be cheap, reliable and power efficient
   Model answer

   An embedded system carries out one dedicated task that never changes, such as controlling a washing machine, so it needs no more power than that task demands. This makes it cheap, reliable and power efficient.

Round 3: Big Think

1. • The address in the PC is copied into the MAR
   • The PC is incremented
   • The address is sent along the address bus
   • The instruction returns along the data bus into the MDR
   • The control unit works out what the instruction means
   • The ALU carries out the addition
   • The result is stored in the accumulator
   • The cycle repeats for the next instruction
   Model answer

   Fetch: the address in the program counter is copied into the MAR, and the PC is incremented. The address is sent along the address bus, and the instruction returns along the data bus into the MDR. Decode: the control unit works out what the instruction means and what it applies to. Execute: the ALU adds the number to the value in the accumulator, and the result is stored in the accumulator. The cycle then repeats for the next instruction.

2. • A higher clock speed completes more instructions per second
   • But it brings more heat and power use
   • More cores execute instructions simultaneously
   • But only if the software splits its work
   • Cache holds frequently used data close to the processor
   • So the processor waits for RAM less often
   • Applied to a game engine, such as physics on separate threads
   • A justified conclusion
   Model answer

   A higher clock speed means more pulses per second, so more instructions are completed every second, but it produces more heat and uses more power, which limits how far it can rise. More cores let several instructions execute simultaneously, which helps a game engine that runs physics, sound and graphics on separate threads, but only if the software is written to split its work; sequential tasks gain nothing. A larger cache holds more frequently used data and instructions close to the processor, so it waits for RAM less often. Overall, for a modern game written to use several threads, more cores and a larger cache are likely to give the bigger gain, so I would choose that processor.

This or That

1. MAR The names tell you: memory address register, memory data register.
2. The ALU The control unit decodes and coordinates. The ALU calculates.
3. It was rewired Holding the program in memory removed the need for rewiring.
4. No: it cuts waiting Cache reduces the time spent waiting for RAM, so more instructions finish each second.
5. A pacemaker A pacemaker does one dedicated task. A laptop is general purpose.
6. During fetch It is incremented as soon as its address has been copied to the MAR.
7. The program counter A jump writes a new address into the PC, so a different instruction is fetched next.
8. In the same memory Sharing one memory is the stored program concept, and the cause of the bottleneck.

Memory Match

• A 2 Program counter: Address of the next instruction
• B 5 MAR: Address being accessed now
• C 6 MDR: Data moving to or from memory
• D 3 Accumulator: Result of the last calculation
• E 4 Control unit: Decodes and coordinates
• F 1 Cache: Fast store for frequently used data

Sort It

1. Execute The ALU adds two numbers Carrying out the instruction is executing.
2. Fetch The PC is incremented It happens during the fetch, not at the end.
3. Fetch The instruction arrives in the MDR It returns along the data bus during the fetch.
4. Execute The ALU compares two values Comparisons happen during execute.
5. Execute The result is stored in the accumulator The result of the execute stage.
6. Execute A jump writes a new address into the PC A jump is executed.
7. Fetch The address in the PC is copied into the MAR The first step of the fetch.
8. Decode The control unit works out what data the instruction needs Working out what it applies to is part of decoding.
9. Decode The control unit works out what the instruction means That is decoding.
10. Fetch The address goes out on the address bus Fetching the instruction from memory.