The Pizza Shop's Latency/Throughput Tradeoff
A friend researching network optimization recently tried to explain the latency/throughput tradeoff to me, but it seemed very abstract and hard to wrap my head around.
It seemed much easier to me to put it in terms of something concrete, something I could understand easily, like my old friend Sungho and his pizza shop.
Sungho is from Busan and used to run a small pizza shop. It was a busy shop with one oven, limiting the number of pizzas that the shop could make per hour. Customers want their pizzas fast and don’t necessarily prepare in advance. When the American expat around the corner calls in his pepperoni pizza (hold the corn, mind you), he wants it as soon as possible. Let’s call the turnaround time on a pizza order latency.
Now, what happens when the frequency and magnitude of orders starts to vary? One might expect more orders during the Chuseok (harvest season) holiday, when families are getting together and there might be a spike of particularly large orders for a whole family. Suppose a Korean family orders 50 pizzas at once. Sungho fires up the oven and starts to churn out pizzas as fast as possible (maximizing throughput), which is what the family wants.
But what then happens, if the lonely expat American calls in immediately after and orders one pepperoni pizza? His latency will be increased to the time it takes to make the 50 pizzas and then his own order.
Sungho might be better served by reducing throughput slightly, allowing some flexibility to minimize the latency of all orders.
More on the Latency Throughput Tradeoff:
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