Validity of the single processor approach to achieving large scale computing capabilities

By G. Amdahl. 1967

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In ML Systems & Hardware

Citations in the corpus

Sourcebook of parallel computing (2003, unknown): “This observation is known as Amdahl’s Law [30].”
Book Reviews: Data-Intensive Text Processing with MapReduce by Jimmy Lin and Chris Dyer (2010, background): “To the extent that the sub-problems are independent [5], they can be tackled in parallel by different workers—threads in a processor core, cores in a multi-core processor, multiple processors in a machine, or many machines in a cluster.”
Bayesian State-Space Modelling on High-Performance Hardware Using LibBi (2013, unknown): “This imposes a serial limitation on the maximum speedup achievable by parallelization with multiple chains, which may be quantified by Amdahl’s law (Amdahl 1967) as Pθ/Bθ.”
Modules for Experiments in Stellar Astrophysics (MESA): Pulsating Variable Stars, Rotation, Convective Boundaries, and Energy Conservation (2019, background): “The theoretical speed up, 1 (1− p) + (p/s) , (46) is predicted by Amdahl (1967), where, for our purposes, s is the number of parallel threads and p is the fraction of the code that will benefit from parallel execution for a given test case.”
Distributed Autonomous Online Learning: Regrets and Intrinsic Privacy-Preserving Properties (2010, background): “and they in turn compute the loss and its gradient on the part of the data they own and communicate it back to the master. However, as [1] observe, when the number of processors increases Amdahls law [2] kicks in and the cost of communicating and synchronizing becomes prohibitively expensive. Anothe”
Developments in the MPI‐M Earth System Model version 1.2 (MPI‐ESM1.2) and Its Response to Increasing CO2 (2019, background): “3 model have substantially improved the scalability of the model to a larger number of processors (Amdahl, 1967), such that longer and/or higher-resolution experiments can now be conducted.”
Complex Concurrent Engineering and the Design Structure Matrix Method (2003, unknown): “According to Amdahl’s law [2], the system’s performance (i.”
Jockey: guaranteed job latency in data parallel clusters (2012, methodology): “Rather than using the simulator above, we can use amodied version of Amdahl’s Law [1] to estimate the job’s completion time given a particular allocation.”
Bundle Methods for Regularized Risk Minimization (2010, methodology): “SpeedupSp = T1 Tp where p is the number of processors andTq is the runtime of the parallelized algorithm onq processors. it; the speedup of overall computation is approaching a limit12 as well-explained by Amdahl’s law (Amdahl, 1967).”
Tackling Exascale Software Challenges in Molecular Dynamics Simulations with GROMACS (2015, background): “The need to design with multi-core and multi-node parallelization in mind at all points is here to stay, which also means Amdahl’s law [3] is more relevant than ever.”