Bandwidth versus Throughput as a Pizza Shop
In a continuation of a talk amongst friends about the throughput versus latency tradeoff, we got confused about the difference between bandwidth and throughput.
While both are rates (quantity over time), bandwidth is the maximum rate while throughput is the rate actually used.
To continue the pizza shop analogy, if I have one oven that fits one pizza can cook a maxmium of ten pizzas per hour, then my bandwidth is ten pizzas per hour. But I don’t have to cook ten pizzas per hour. If I leave the oven off, I can cook zero pizzas per hour. Or maybe I only want to cook five pizzas per hour. The rate at which I actually cook pizzas is my throughput.
How does this relate to the latency versus throughput tradeoff? And why is it a tradeoff between latency and throughput and not between latency and bandwidth? If your throughput is maxed out at your bandwidth, then latency is likely to be high. If your bandwidth is high, you can serve at a higher rate, but your latency still depends on your throughput being lower than your bandwidth.
More
Archive
chinese tang-dynasty-poetry 李白 python 王维 rl pytorch numpy emacs 杜牧 spinningup networking deep-learning 贺知章 白居易 王昌龄 杜甫 李商隐 tips reinforcement-learning macports jekyll 骆宾王 贾岛 孟浩然 xcode time-series terminal regression rails productivity pandas math macosx lesson-plan helicopters flying fastai conceptual-learning command-line bro 黄巢 韦应物 陈子昂 王翰 王之涣 柳宗元 杜秋娘 李绅 张继 孟郊 刘禹锡 元稹 youtube visdom system sungho stylelint stripe softmax siri sgd scipy scikit-learn scikit safari research qtran qoe qmix pyhton poetry pedagogy papers paper-review optimization openssl openmpi nyc node neural-net multiprocessing mpi morl ml mdp marl mandarin macos machine-learning latex language-learning khan-academy jupyter-notebooks ios-programming intuition homebrew hacking google-cloud github flashcards faker docker dme deepmind dec-pomdp data-wrangling craftsman congestion-control coding books book-review atari anki analogy 3brown1blue 2fa