Skip to content
  • Categories
  • Recent
  • Tags
  • Popular
  • World
  • Users
  • Groups
Skins
  • Light
  • Cerulean
  • Cosmo
  • Flatly
  • Journal
  • Litera
  • Lumen
  • Lux
  • Materia
  • Minty
  • Morph
  • Pulse
  • Sandstone
  • Simplex
  • Sketchy
  • Spacelab
  • United
  • Yeti
  • Zephyr
  • Dark
  • Cyborg
  • Darkly
  • Quartz
  • Slate
  • Solar
  • Superhero
  • Vapor

  • Default (No Skin)
  • No Skin
Collapse
Code Project
  1. Home
  2. General Programming
  3. Hardware & Devices
  4. RAM Performance

RAM Performance

Scheduled Pinned Locked Moved Hardware & Devices
2 Posts 2 Posters 0 Views 1 Watching
  • Oldest to Newest
  • Newest to Oldest
  • Most Votes
Reply
  • Reply as topic
Log in to reply
This topic has been deleted. Only users with topic management privileges can see it.
  • T Offline
    T Offline
    t mhill
    wrote on last edited by
    #1

    I am in the middle of writing a program that calculates the performance of different algorithsm. Im trying to implement a class that can estimate the peformance before any calculations have taken place. Obviously RAM/CPU speed will have a very big influence, so i have taken numerous calculations using machines of different specifications, however i need to fit a Trend Line to it. Does anyone know an 'estimate' to RAM/CPU Performance i.e Log, Exponential, linear? Thanks MH

    A 1 Reply Last reply
    0
    • T t mhill

      I am in the middle of writing a program that calculates the performance of different algorithsm. Im trying to implement a class that can estimate the peformance before any calculations have taken place. Obviously RAM/CPU speed will have a very big influence, so i have taken numerous calculations using machines of different specifications, however i need to fit a Trend Line to it. Does anyone know an 'estimate' to RAM/CPU Performance i.e Log, Exponential, linear? Thanks MH

      A Offline
      A Offline
      Alexander M
      wrote on last edited by
      #2

      Measuring CPU is easier than measuring ram. You create an algorithm without heavy memory access and a loop calling it. now you can either set a constant loop count and measure the time to execute, or you create a multithreaded application and interrupt the loop after a fixed timeout. RAM is more difficult, as you need to access it using CPU instructions... That means, that the score will strongly depend on the kind of algorithm you use to measure the performance! I'd suggest using SSE instructions, since you can bypass the CPU cache that way. MOVNTQ for example. Most benchmarks I've seen use a linear scale to measure both CPU and RAM. But that also depends on how to evaluate the score with you application.

      Don't try it, just do it! ;-)

      1 Reply Last reply
      0
      Reply
      • Reply as topic
      Log in to reply
      • Oldest to Newest
      • Newest to Oldest
      • Most Votes


      • Login

      • Don't have an account? Register

      • Login or register to search.
      • First post
        Last post
      0
      • Categories
      • Recent
      • Tags
      • Popular
      • World
      • Users
      • Groups