Amdahl's law
en.wikipedia.org · 3,059 words · saved by 9 readers
Couldn't find lead section for Amdahl's_law
Amdahl's law - Wikipedia Jump to content From Wikipedia, the free encyclopedia Formula in computer architecture Amdahl's law demonstrates the theoretical maximum speedup of an overall system and the concept of diminishing returns. Plotted here is logarithmic parallelization vs linear speedup. If exactly 50% of the work can be parallelized, the best possible speedup is 2 times. If 95% of the work can be parallelized, the best possible speedup is 20 times. According to the law, even with an infinite number of processors, the speedup is constrained by the unparallelizable portion. In computer arc
saved by
- Anson Yu
- HudZah
- Aryan Naik
- Karthik Suresh
- Stephen Fay
- Lydia Nottingham
- Vincent Cheng
- Alex Wang
- SUJASH AGRAWAL
related reading
- Scaling Laws, Carefully | Lil'Loglilianweng.github.io
- Wirth's lawen.wikipedia.org
- Engelbart's law - Wikipediaen.wikipedia.org
- Transformer Inference Arithmetic | kipply's blogkipp.ly
- All About Rooflines | How To Scale Your Modeljax-ml.github.io
- Zipf's law - Wikipediaen.wikipedia.org
- Modern Microprocessors - A 90-Minute Guide!lighterra.com
- Memory access is O(N^[1/3])vitalik.eth.limo
- Metcalfe's law - Wikipediaen.wikipedia.org
- abseil / Performance Hintsabseil.io
- Embarrassingly parallel - Wikipediaen.wikipedia.org
- Power law - Wikipediaen.wikipedia.org