AP CSP Topic 4.3 Guided Notes - Parallel and Distributed Computing
Big Idea 4: Computer Systems and Networks · Topic 4.3 · Guided Notes (Student)
Parallel and Distributed Computing — Guided Notes
Fill these in during class or catch up here if you were absent. Print this page or work on paper — then check yourself with the CFUs on the Topic 4.3 page.
Print these notesTopic 4.3 lesson pageAll CSP topics
Today’s objectives
- Compare sequential, parallel, and distributed solutions — and tell the last two apart by counting devices (LO CSN-2.A)
- Compute totals and speedup: sum for sequential, sequential-plus-longest-parallel, sequential ÷ parallel (LO CSN-2.A)
- Describe the benefits and limits of parallel and distributed computing (LO CSN-2.B)
Bell ringer
The class must wash 60 dishes after the lab breakfast. One student working alone takes 60 minutes: 45 minutes of scrubbing plus 15 minutes loading the single drying rack, which only one person can use at a time. Today THREE students share the job — the scrubbing splits evenly, but the rack line still moves one person at a time.
Predict first: is the answer 20 minutes? Now compute the real finish time. Then find the time that NO number of extra washers could ever beat.
01. Three Ways to Compute
Key Vocabulary (LO CSN-2.A)
| Term | Definition (write it) |
|---|---|
| Sequential computing | |
| Parallel computing | |
| Distributed computing | |
| Speedup |
Same Job, Three Models
- The sequential model's defining rule: nothing starts until .
- Parallel computing saves time because some of the smaller operations run .
- To compare the efficiency of two solutions, compare .
Parallel vs. Distributed — The Classic Mix-Up
Parallel computing
Distributed computing
- Parallel computing splits one program across the cores of .
- You know it's DISTRIBUTED (not parallel) when the program runs across .
- The fastest way to tell the two models apart on the exam is to count .
AP TIP: Count the devices. One machine, many cores = PARALLEL. Many machines, one program = DISTRIBUTED.
02. Do the Math: Time and Speedup
Do the Math: Total Time and Speedup
The CED's model: sequential time = SUM of all steps; parallel time = sequential tasks + LONGEST parallel task; speedup = sequential ÷ parallel.
Complete the empty cells.
| Task set (each parallel task gets its own processor) | Sequential time (sum of all steps) | Parallel time (sequential tasks + longest parallel task) | Speedup (sequential ÷ parallel) |
|---|---|---|---|
| Two parallel tasks: 30 s and 50 s (no sequential task) | 30 + 50 = 80 s | 50 s (longest parallel task; nothing sequential to add) | 80 ÷ 50 = 1.6 |
| Sequential task: 20 s, then parallel tasks: 40 s and 60 s | |||
| Three parallel tasks: 30 s, 60 s, 30 s (no sequential task) | 30 + 60 + 30 = 120 s | ||
| Sequential task: 60 s, then parallel tasks: 20 s, 20 s, 20 s | 60 + 20 + 20 + 20 = 120 s |
Stop and think
- A program has one 15-second sequential task, then two parallel tasks of 25 seconds and 35 seconds. Compute the sequential time, the parallel time, and the speedup. Show the arithmetic.
- In the table's last row, THREE processors produced a speedup of only 1.5. Name the number in that task set that kept the speedup low, and state what would have to change for the speedup to rise.
- A tablet applies one video filter using all 8 cores of its chip. A weather service runs one forecast model across 2,000 separate computers. Classify each as parallel or distributed and name the one question that decides it.
Show your work and answer in complete sentences. Then check yourself with the matching CFUs on the Topic 4.3 page.
03. Benefits, Limits, and When Distributed Wins
Watch out — “Doubling the Processors Halves the Time”
Myth: If 2 processors make a program twice as fast, then 4 processors will make it four times as fast, and 64 will make it sixty-four times as fast — speedup simply tracks the processor count.
Explain why this is wrong:
Deep Dive · Beyond the AP Exam
Deep Dive: The Speedup Ceiling (Amdahl's Law)
Push the washers to infinity
Amdahl's idea, one formula
- In the dish job, no number of washers can ever finish faster than .
- The maximum possible speedup equals the total sequential time divided by .
- When processor-buying stops paying off, the engineering fix is to shrink .
Deep dive only — the AP exam never asks for Amdahl's Law by name; it tests the idea behind it: the sequential portion sets the ceiling.
Stop and think
- A program spends 30 seconds in its sequential portion and 90 seconds in work that can be split into as many parallel tasks as you like. What total time does the solution approach as processors are added, and what maximum speedup does that put a ceiling at? Show the division.
- A hospital network's imaging archive is too large for any single computer to store, and one machine would need decades to analyze it. Name the computing model that solves this, and the TWO single-computer limits the CED says it overcomes.
- A classmate says: 'Parallel computing scales exactly like sequential computing — both just get faster when you buy better hardware.' Correct the claim using the CED's comparison of how the two models scale.
Show your work and answer in complete sentences. Then check yourself with the matching CFUs on the Topic 4.3 page.
Common AP Traps
Three ways Topic 4.3 loses points on the exam — one minute now, real points in May.
The forgotten sequential task — in your own words:
Speedup is a ratio — in your own words:
Parallel ≠ distributed — in your own words:
Parallel and Distributed Computing, in One Slide
- Sequential = operations in order, one at a time; parallel = one program broken into smaller sequential operations, some simultaneous; distributed = multiple devices running one program.
- Compare efficiency by comparing the time solutions take to perform the same task.
- Sequential time = the SUM of all steps; parallel time = the sequential tasks + the LONGEST parallel task.
- Speedup = sequential time ÷ parallel time — a ratio, never a subtraction.
- Parallel scales better than sequential but is still limited by its sequential portion — at some point more processors stop helping; distributed computing solves larger problems faster, and problems no single computer could solve at all.
Exit check — I can…
- ☐ tell sequential, parallel, and distributed apart by counting devices (LO CSN-2.A)
- ☐ compute totals: sum for sequential; sequential + longest parallel task (LO CSN-2.A)
- ☐ compute speedup as sequential time ÷ parallel time (LO CSN-2.A)
- ☐ explain the sequential-portion ceiling and when only distributed works (LO CSN-2.B)
Get in Touch
Whether you're a student, parent, or teacher — I'd love to hear from you.
Just want free AP CS resources?
Enter your email below and check the subscribe box — no message needed. Students get daily practice questions and study tips. Teachers get curriculum resources and teaching strategies.
Message Sent!
Thanks for reaching out. I'll get back to you within 24 hours.
Prefer email? Reach me directly at [email protected]