latency   3392

« earlier    

What’s Better Than 40 GPU-based Computers? A Computer With 40 GPUs
Engineers aim to use “silicon interconnect fabric” to build a computer with 40 GPUs on a single silicon wafer
wafer  scale  integration  latency  trilogy  systems  IEEE 
7 days ago by Bluyssae
Little’s Law- An insight on the relation between latency and throughput | Flux7
> The basic relationship between latency and throughput within a stable system—one that is not ramping up and down—is defined by a simple equation called Little’s Law:
> occupancy = latency x throughput
latency  littleslaw 
9 days ago by nwjsmith
Cake -
Seemingly the currently most popular answer to reducing bufferbloat-induced latency
networking  bufferbloat  queueing  latency  router 
10 days ago by michielbuddingh
Amazon: Why do we measure latency in percentiles?
If you're wondering what "P-four-nines" means, it's the latency at the 99.99th percentile, meaning only one in 10,000 requests has a worse latency. Why do we measure latency in percentiles?

A thread about how how it came to be at Amazon...
performance  latency  statistics  webeng  webengineering  cloudcomputing  amazon  aws 
18 days ago by jevon

« earlier    

related tags

10gbe  10gbit  2017  2018  2019  5g  adorjan  algorithmic  algorithms  amazon  applications  apps  architecture  argument  atom  audio  autonomous_vehicles  aws  azure  bandwith  benchmark  benchmarking  benchmarks  betting  big-o  bound  bsd  bufferbloat  c-ran  cache  chip  cli  cloud  cloudcomputing  cloudflare  co-viewing  comparison  complexity  computer  computers  computing  control  controller  cool  cores  correction  costs  coteditor  cpu  cqrs  cs  cycles  danluu  data  datacenter  debugging  design  development  devops  display  distributed-systems  distributed  driver  dynamodb  e-ink  editors  elgato  encoder  encoding  engineering  esports  everyprogrammershouldknow  exmachinagtoup  facebook  firmware  fps  game  games  gamify  gaming  go  goetz  golang  google  hack  haivision  hardware  hbase  history  hn  hollywood  howto  http  human  ieee  immersive  imp  industrial_internet  inertia  injection  input  integration  io  iolatency  iomax  ioping  ios  java  keyboard  keyboards  kinesis  knowledge  lag  lambda  learning  linux  littleslaw  logging  loopback  low  mac  macos  management  manufacturers  matt  maximum  measurement  measurements  mec  memory  metadata  metrics  microservices  midi  milicast  millicast  mobile_phones  modem  netbanking  network  networking  nginx  nike_station  notation  notepad  numa  number  numbers  nvm  nvme  of  optimal  optimization  orleans  osx  ovum  p9999  paas  pagespeed  paper  per  perf  performance-tuning  performance  phenix  piperesearch  precision  profiling  programming  python  qos  queueing  quiz  realtime  regions  replication  resources  responsiveness  robotics  route53  router  rps  rtt  rx  scalability  scale  scaling  scheduler  schedulers  serverless  service-ownership  service  sharing  shell  shop  simulation  size  slam_line  slides  smartphones  software  speed  sports  sportscasting  srt  stapelberg  statistics  stats  storage  streaming  stun  surveillance  sysadmin  systems  tail  tcp  telemedicine  telephony  tempus-fugit  terminal-emulator  terminal  test  testing  theory  threads  throughput  time  tool  tools  tootme  toread  tp95  transcoder  trilogy  tti  turn  twitter  ui  usbmic  verizon  video  videogames  virtual_reality  virtualization  wafer  waveshare  web  webeng  webengineering  webrtc  window  youtube  zones 

Copy this bookmark: