← Back to examples
One Continuous Idea: thread → data center
The whole course was never fifteen separate topics — it was one idea,
zoomed out step by step. Click each scale to see the same "divide, run at once, combine" pattern,
the tool you'd use there, and which session covered it.
☜ Click any box above · they're all the same idea at a bigger radius
The pattern that never changes: split the work into independent pieces, run them at once,
keep coordination cheap, and combine the results. The tools change with scale; the idea does not.
… and the five units that got you here
Unit I — Why + Classify + Models: free lunch over; concurrency vs parallelism; Flynn's SISD/SIMD/MISD/MIMD; shared vs distributed memory; threads & fork-join. (S01–03)
Unit II — Architecture: interconnects (bus/mesh/torus/hypercube); memory hierarchy, cache coherence (MESI), false sharing; CPUs vs GPUs, SIMT. (S04–06)
Unit III — Design + Measure: Foster's PCAM; Amdahl's Law & Gustafson; speedup & efficiency; load balancing & work stealing. (S07–09)
Unit IV — Tools: OpenMP (shared, one node), MPI (distributed, many nodes), CUDA (the GPU). Labs 1–5. (S10–12)
Unit V — Applications: scientific computing & parallel databases; big data (MapReduce, Spark); real-time systems & the AI frontier. (S13–15)