Course contentsShow
← Computer Science
Lesson 1173 of 2,87227. Pipelining and PerformanceFree lesson

Control Hazard Mitigation Strategies

Overview of techniques including branch prediction, delayed branches, and speculative execution to reduce control penalties.

Control Hazard Mitigation Strategies

What you'll learn: How processors reduce performance penalties caused by branch instructions in pipelined execution.

The Problem

You already know that control hazards occur when the pipeline encounters a branch instruction and doesn't know which instruction to fetch next until the branch resolves. Waiting for the answer creates costly pipeline stalls. But processors don't just accept this penalty—they fight back with clever mitigation strategies.

Key Mitigation Techniques

Branch Prediction

Instead of waiting, the processor guesses which direction the branch will take (taken or not taken) and continues fetching instructions along that path. If the guess is correct, no time is lost! If wrong, the pipeline flushes the incorrectly-fetched instructions and recovers.

Analogy: Like betting on which route home will be faster during rush hour. If you're right, you save time. If wrong, you backtrack—but often the gamble pays off.

Delayed Branches

The compiler rearranges code to place useful instructions immediately after the branch—instructions that execute regardless of the branch outcome. These "delay slots" keep the pipeline productive while the branch resolves.

Analogy: While waiting at a traffic light, you adjust your mirror or radio—small tasks that need doing anyway.

Speculative Execution

The processor doesn't just fetch speculatively; it executes the predicted path's instructions tentatively. Results are held temporarily and only committed if the prediction was correct. This maximizes work done during the branch resolution window.

Why This Matters

These strategies transform control hazards from pipeline-killers into manageable penalties. Modern processors combine all three techniques to maintain high instruction throughput even in code with frequent branches.

Key Takeaway: Control hazard mitigation uses branch prediction to guess outcomes, delayed branches to fill waiting time productively, and speculative execution to perform tentative work—all aiming to minimize pipeline stalls caused by branch instructions.