ConceptSystems & CS
Instruction-level parallelism
Processor hardware exploits instruction-level parallelism by finding opportunities to execute multiple instructions simultaneously in different pipeline stages.
Know first
Named as a prerequisite by the lectures that define this. The number is how many say so. A grey one has no page yet.
Hardware abstraction1Instruction1Load and store instructions1Parallel computing1Pipelined instruction execution1Processor pipelining1Program dependencies1
Opens up
Concepts that name this one as a prerequisite.
Lecturers
3Each in their own terms, from the easiest level.
Charles Leiserson
Achieving higher performance by reordering instructions to cover load latency.Kayvon Fatahalian & Kunle Olukotun
Processors did in fact leverage parallel execution to make programs run faster, it was just invisible to the programmer.
Lectures that assume it
4- Problem Solving V: Digital Design and Computer Architecture (ETH Zürich, Spring 2025, Lecture 30)
- Dataflow, Superscalar Execution & Branch Prediction
- VLIW and Systolic Array Architectures
- Prefetching II, Processing-in-Memory, and Parting Thoughts (ETH Zürich: Digital Design and Computer Architecture, Lecture 25)
Connections
12The lectures that define it or assume it, and the concepts next to it.
Concepts5
Lectures7
- Assembly Language & Computer Architecture
- Synchronization Without Locks
- Parallel Computing: Why Parallelism? Why Efficiency?
- Problem Solving V: Digital Design and Computer Architecture (ETH Zürich, Spring 2025, Lecture 30)
- Dataflow, Superscalar Execution & Branch Prediction
- VLIW and Systolic Array Architectures