Video lesson: Branch Prediction and Exposing Parallel Work

Lesson promise

By the end, the learner should be able to explain the core model for branch prediction and exposing parallel work, apply it to a concrete input, and identify when its usual shortcut or guarantee stops applying. This is a recording brief; publish it as a playable lesson after the narration and visual sequence have been produced and reviewed.

Narration draft

Branches make program control depend on data. A modern CPU predicts likely directions so it can keep fetching instructions before the condition is fully resolved. Correct predictions preserve pipeline flow; incorrect ones require discarding speculative work and refilling parts of the pipeline.

A loop over sorted values may have a highly predictable branch, while a branch on randomized values may be difficult to predict. Some code can be expressed with conditional operations or grouped data to reduce unpredictable control flow, but those rewrites can increase instruction count or complicate correctness.

Instruction-level parallelism is limited by dependencies, branches, and memory stalls. Removing a branch is not automatically faster; the new operations may cost more or prevent vectorization. Benchmark realistic input distributions and check that the compiler generated the intended code.

Visual sequence

1. Put the input and assumptions on screen. Ask the learner to predict the next state before revealing it.
2. Animate the representation and show the operation one transition at a time.
3. Pause at the boundary case in the companion article and compare the result with the invariant.
4. End with the exercise prompt: You have a filter with a branch that is true for almost every element. Explain why that may perform differently from a branch that is true half the time, and name one measurement you would collect.

Companion material

Use the article, trace, and interactive concept flow as the learner’s written and visual references. The video remains planned until an actual playable media URL and reviewed transcript are available.