Linked from

The 28 pages that link to Parallel computing, each with the reason it gives.

• Graphics processing unitNarrower topic: GPUs are specialized hardware for exploiting large-scale parallelism.

• Moore's lawCompared with: It is a performance strategy, not a prediction about chip transistor density.

• Distributed computingCompared with: Parallelism can occur within one tightly coupled machine; distribution requires networked coordination.

• High-performance computingNarrower topic: HPC gains speed by distributing suitable workloads across many processors.

• Scientific computingRelated: Parallelism lets large simulations finish within practical time and resource limits.

• AssociativityRelated: Associative reductions can combine partial results in parallel.

• Amdahl's lawNarrower topic: Amdahl's law is commonly used to estimate its scaling limits.

• Functional programmingRelated: Independent computations over immutable values can often run without coordination over shared state.

• SupercomputerNarrower topic: Supercomputers gain speed by splitting calculations into tasks that can run concurrently.

• SynchronizationCompared with: Simultaneous execution can occur without requiring all operations to stay in step.

• Von Neumann architectureCompared with: It contrasts with the simple sequential execution model often associated with the architecture.

• CUDANarrower topic: CUDA is one platform for expressing and executing parallel computations.

• Concurrency (computer science)Compared with: Concurrency permits overlapping progress without requiring simultaneous execution.

• Declarative programmingRelated: When dependencies are stated rather than sequenced, execution systems may identify parallel work.

• Parallel algorithmNarrower topic: Parallel algorithms are the software methods used within this broader computing approach.

• Adaptive mesh refinementRelated: AMR requires refined patches to be distributed and rebalanced across processors.

• Gustafson's lawNarrower topic: The law addresses how performance in this broader field changes with processor count.

• Computational scienceRelated: Parallelism makes large scientific calculations feasible within practical time.

• Gene AmdahlNarrower topic: Amdahl’s law became a standard way to reason about the limits of parallel systems.

• Hardware accelerationNarrower topic: Parallelism is a main source of speedup in accelerator workloads.

• Concurrent computingCompared with: Parallelism requires simultaneous execution; concurrency only requires overlapping progress.

• MultiprocessingNarrower topic: Multiprocessing is one way to perform parallel computing.

• Seymour CrayCompared with: Cray’s celebrated vector designs differed from systems built primarily around many processors working together.

• Computational astrophysicsRelated: Large astrophysical models divide calculations across many processors.

• Computer clusterRelated: Many clusters gain speed by dividing parallelizable work among nodes.

• Computer multitaskingCompared with: Unlike task switching on one processor, parallel execution requires simultaneous computational capacity.

• General-purpose computing on graphics processing unitsNarrower topic: GPGPU is a hardware-specific approach to parallel computing.

• Multi-core processorNarrower topic: Multi-core processors provide parallel hardware, but software must expose suitable work.