Validity of the Single Processor Approach to Achieving Large Scale Computing Capabilities.¶
Amdahl, G. M. (1967). Validity of the Single Processor Approach to Achieving Large Scale Computing Capabilities. Proceedings of the April 18–20, 1967 Spring Joint Computer Conference, 30, 18-20.
Cited by¶
13 citations across 13 artifacts.
Each citation links to the sentence it supports in the citing article.
Primes¶
- Bottleneck
- Amdahl's law formalizes the ceiling: the maximum speedup from parallelizing a program is bounded by the fraction that remains inherently serial, so the serial fraction is the computational bottleneck no amount of added cores can overcome.
This sourceAFIPS. Establishes Amdahl's law: the maximum speedup from parallelization is bounded by the inherently serial fraction, making that fraction the bottleneck on speedup.
- Amdahl's law formalizes the ceiling: the maximum speedup from parallelizing a program is bounded by the fraction that remains inherently serial, so the serial fraction is the computational bottleneck no amount of added cores can overcome.
- Concurrency
- Speed-up potential versus coordination overhead trade-off
This sourceSUPPORTS marker 051 (speed-up potential vs coordination overhead trade-off). DOI verified.
- Speed-up potential versus coordination overhead trade-off
- Irreducible Floor
- Manufacturing quality: a given process variance produces a defect-rate floor, and pushing for "zero defects" inside that process either inflates inspection cost or destabilises the process; lowering the floor requires process redesign, not more inspection. Network and software performance: irreducible latency (physical propagation plus minimum processing) and Amdahl's-law speedup limits (the serial fraction bounds achievable speedup regardless of core count) are floors no intra-regime tuning can pierce.
This sourceStates Amdahl's law: the serial fraction s bounds achievable speedup at 1/s regardless of processor count, setting an execution-time floor.
- Manufacturing quality: a given process variance produces a defect-rate floor, and pushing for "zero defects" inside that process either inflates inspection cost or destabilises the process; lowering the floor requires process redesign, not more inspection. Network and software performance: irreducible latency (physical propagation plus minimum processing) and Amdahl's-law speedup limits (the serial fraction bounds achievable speedup regardless of core count) are floors no intra-regime tuning can pierce.
- Premature Optimization
- Scalability
- The characterization of the scaling relation: linear (ideal), sublinear (diminishing returns), logarithmic, bounded (Amdahl-limited by serial fraction), or degrading (anti-scalable due to coordination overhead)
Supported in partVerified against the publisher's abstract
Amdahl (1967) supplies at most the serial-fraction bound behind the claim's 'bounded' class; neither its abstract nor anything else we hold states the five-way scaling-relation taxonomy.
“For over a decade prophets have voiced the contention that the organization of a single computer has reached its limits and that truly significant advances can be made only by interconnection of a multiplicity of computers in such a manner as to permit cooperative solution.”
- The characterization of the scaling relation: linear (ideal), sublinear (diminishing returns), logarithmic, bounded (Amdahl-limited by serial fraction), or degrading (anti-scalable due to coordination overhead)
- Scale
Domain-specific¶
- Amdahl's Law
- Gene Amdahl's 1967 paper made the underlying argument while challenging projections for large-scale parallel organization; later performance literature commonly presents the normalized equation above as Amdahl's Law.
This sourceFoundational primary source; bibliographic identity verified through ACM.
- Gene Amdahl's 1967 paper made the underlying argument while challenging projections for large-scale parallel organization; later performance literature commonly presents the normalized equation above as Amdahl's Law.
- Sun–Ni Law
Mechanisms¶
- Benchmark Harness
- Its central trap is the microbenchmark that does not reflect the whole system: a dramatic local speed-up on a routine that is a tiny fraction of total runtime barely moves anything
This sourceEstablishes that accelerating a small fraction of a computation yields only a limited improvement in overall performance.
- Its central trap is the microbenchmark that does not reflect the whole system: a dramatic local speed-up on a routine that is a tiny fraction of total runtime barely moves anything
- Horizontal Scale-Out
- Adding units does nothing for a task bound by a single serialized step or a shared resource every unit must queue for — throughput is capped by the fraction of the work that must run in sequence, so scaling out the parallel part yields ever-smaller returns while the serial part dominates.
This sourceAmdahl shows that a fixed serial fraction caps total speedup, producing diminishing returns as more resources accelerate only the parallel portion.
- Adding units does nothing for a task bound by a single serialized step or a shared resource every unit must queue for — throughput is capped by the fraction of the work that must run in sequence, so scaling out the parallel part yields ever-smaller returns while the serial part dominates.
- Parallel Server Activation
- Parallel speedup is bounded by whatever remains serial: past a point, more instances just heap load onto the one shared database and throughput flattens or degrades (Amdahl's Law
This sourceShows that a serial fraction imposes an upper bound on parallel speedup, so returns eventually flatten.
- Parallel speedup is bounded by whatever remains serial: past a point, more instances just heap load onto the one shared database and throughput flattens or degrades (Amdahl's Law
- Representative Workload Profiling
- Its strength is that it directs scarce refinement effort by evidence, and it is the natural home for Amdahl's law
This sourceShows that whole-system improvement is limited by the fraction of total execution affected by an enhancement.
- Its strength is that it directs scarce refinement effort by evidence, and it is the natural home for Amdahl's law
- Scale-Sweep Benchmark
Verification¶
Does it exist? Confirmed. This work's DOI resolves to a registered record, which fixes its identity. That is all it fixes.
Does it back the claim? Read against the text for 1 of 13 citations: 1 supported in part. Each verdict is shown under its citation below, with what in the work backs the sentence.
Support is checked per citation rather than per work — the same source can be cited soundly in one article and wrongly in another. Per-citation recording began recently, so a citation with no recorded check is a gap in the record rather than evidence it went unchecked.
See how references were verified.
Registry ID ref:2ae8547511ef · see in the full table