cooling options

entropyentropy Yah-Der-Hey (Wisconsin)
edited May 2004 in Hardware
i think i'm gonna get that $20 slk-947 instead of the sp-94 as the differences in cooling really aren't all that great, along with a smartfan 2. which is better - AS5 or Ceramique? i don't really know anything about those so if anyone has any links/comments, thanks. also, for people OC'ing from where i'm at anywhere to 3.4 or so, if you remember, what was the appx. difference you saw in temps going from stock to this combo?

Comments

  • Geeky1Geeky1 University of the Pacific (Stockton, CA, USA)
    edited May 2004
    The thermal compound makes basically no difference. I'd just get the ceramique. I use it, and it does its job. :)

    Can't help you with the temps until I get the 2.8E fixed. :-/
  • entropyentropy Yah-Der-Hey (Wisconsin)
    edited May 2004
    ok thanks. the other question i had, but forgot, was what exactly is the "break in" period or whatever it's called? how long is it, and should i try to keep the processor at what it'll be most of the time (400 mhz overclock, 100% usage dual folding) or should i let it idle for that time??
  • csimoncsimon Acadiana Icrontian
    edited May 2004
    :fold: fold 100% ...the true test!!!
    Monitor your temps.
  • Geeky1Geeky1 University of the Pacific (Stockton, CA, USA)
    edited May 2004
    IMO, the "break in" period is basically a farce. I maintain that the "superior performance" of arctic silver's products is mostly snake oil. I do use Ceramique, and I'm happy with it, but I use it because it's cheap and it *might* be better than regular compound, not because I have any evidence that it is.
  • Straight_ManStraight_Man Geeky, in my own way Naples, FL Icrontian
    edited May 2004
    Major computer OEMs that test reasonably heavily for stability run the CPUs for two-three days doing a variety of benchmarks and tasks, and shoot for minimum of 80% of designed load and Max of 100% with some surges of overload of 5% allowed in that test, continuous run. Extreme boxes get at least that much if the OEM does not want huge return rates. Thats for production, where designs of sets of things have been narrowed down to a few combos that should work and have been run and prove to work.

    So, guesstimate, to prove in just your hardware set, I would run stock for about a week and do on it as much as possible-- a week worth of running hours, at least 24-48 continuous run included in that week with the rest as needed for a home user box. If things are not overstretched and you try to do most of what you want to do in that time, you probably will have minimal failures beyond first week if you use a good PSU and protect it from major surges and protect XP from sudden power drops without a controlled shutdown before your power backup fails.

    Then, if I choose to, I OC a bit just to see how good the set of things I chose is. If it is for a normla user that I am building a box, I will then return the settings to stock just to be able to troubleshoot on the phone. If the user is knowledgeable and an extreme user I will leave things OC'd after testing for longer.

    Burning in is what this is called. MOST things sold for IT are not individually tested real well, 80-90% of things mfr'd for IT use are sample tested. If you get things that are real well tested, and are guaranteed longer, expect to pay a premium for things of that sort from mfrs who have a long track record of doing what they specialize in. so, if on a budget at all, and new to building, test in for minimum of three days to make sure things are up to normal, then OC slowly, would be my advice.

    If you want that short a test, throw everything you can think of to do with that box at it in that time-- take a weekend, and get the box busy as heck all weekend even if it not doing things with you there, and leave the thermal stuff enabled if you will be gone long anytime in there. Then OC AFTER that time. 48-72 hours of runtime are normal burnin theese days for established OEMs, I test minimum that long for ANYTHING I build or buy prebuilt before custormer gets it.

    I am saying things twice or three times for emphasis, and to show how to do a kickass burnin. Google burnin for more.
  • MedlockMedlock Miramar, Florida Member
    edited May 2004
    Arctic Sliver website has instructions too....
    Due to the unique shape and sizes of the particles in Arctic Silver 5's conductive matrix, it will take a up to 200 hours and several thermal cycles to achieve maximum particle to particle thermal conduction and for the heatsink to CPU interface to reach maximum conductivity. (This period will be longer in a system without a fan on the heatsink or with a low speed fan on the heatsink.) On systems measuring actual internal core temperatures via the CPU's internal diode, the measured temperature will often drop 2C to 5C over this "break-in" period. This break-in will occur during the normal use of the computer as long as the computer is turned off from time to time and the interface is allowed to cool to room temperature. Once the break-in is complete, the computer can be left on if desired.
Sign In or Register to comment.