Pipeline Stalls and Bubbles
What You'll Learn
You'll learn how processors insert "bubbles" (NOPs or no-operation instructions) into the pipeline to prevent data hazards from causing incorrect results.
The Problem: Instructions That Can't Proceed
When a data hazard occurs—say, instruction B needs a result that instruction A hasn't finished computing yet—the pipeline has a problem. If B tries to execute too early, it will read wrong data. The solution? Force B to wait.
What is a Pipeline Stall?
A pipeline stall (also called a pipeline bubble) is an intentional delay inserted into the pipeline. The processor stops fetching new instructions and inserts a NOP (no operation) into one or more pipeline stages.
Think of it like a factory assembly line: if Station 3 needs a part that Station 2 hasn't finished yet, Station 3 must idle (do nothing) for one or more cycles until the part is ready.
How Bubbles Work
When the control unit detects a data hazard, it:
- Freezes the instruction that needs data (keeps it in the same stage)
- Inserts a bubble (NOP) into the next stage
- Repeats until the data dependency is resolved
The bubble travels through the pipeline like an empty slot, wasting that cycle but ensuring correctness.
Example Scenario
Cycle 1: ADD R1, R2, R3 (computes R1, finishes in cycle 4)
Cycle 2: SUB R4, R1, R5 (needs R1 — hazard detected!)
→ Insert bubble, stall SUB
Cycle 3: (bubble travels through pipeline)
Cycle 4: SUB can now safely read R1
The bubble costs performance—it's wasted time—but guarantees the SUB instruction reads the correct value from R1.
Key Takeaway: Pipeline bubbles are intentional NOPs inserted to delay dependent instructions, resolving data hazards at the cost of performance by creating idle cycles in the pipeline.