Results 1 to 7 of 7
  1. #1
    4 Star Lounger
    Join Date
    Jun 2010
    Location
    Littleton, Colorado
    Posts
    492
    Thanks
    33
    Thanked 6 Times in 6 Posts

    XP kernel loaded into RAM?

    Windows XP SP3, 3GB RAM

    I recently changed the registry to allow loading the XP kernel totally into RAM (although I didn't change my Pagefile size).
    How can I verify that the kernel is actually running in RAM?

    Thanks.

  2. The Following User Says Thank You to LesF For This Useful Post:

    RangerPHX (2013-05-29)

  3. #2
    Super Moderator Deadeye81's Avatar
    Join Date
    Dec 2009
    Location
    North Carolina, USA
    Posts
    2,654
    Thanks
    7
    Thanked 113 Times in 97 Posts
    Hi LesF,

    You might be interested in reading TweakHounds Super XP Tweaking Guides article on Bad Tweaks. Changing the registry to load the XP kernel into RAM falls within the subject matter of this article.

    There are a number of interesting bits in the article about popular tweaks that are useless or just unnecessary.

  4. #3
    4 Star Lounger
    Join Date
    Jun 2010
    Location
    Littleton, Colorado
    Posts
    492
    Thanks
    33
    Thanked 6 Times in 6 Posts
    Thanks for your reply, Deadeye81. That was a very interesting site. In fact, it specifically mentioned the 3 registry changes I made.
    Which brings up another question. The site I used to identify the tweaks I made says that there is a noticeable performance improvement from them (don't know which XP version). The site you provided says otherwise. What is the best way to actually compare performance with and without the tweaks?

  5. #4
    Super Moderator Deadeye81's Avatar
    Join Date
    Dec 2009
    Location
    North Carolina, USA
    Posts
    2,654
    Thanks
    7
    Thanked 113 Times in 97 Posts
    I do not know of a way to measure the actual performance to test both scenarios. I can only offer my own observation. I used this tweak a few years ago, and found no discernible increase in performance. In my book, if I cannot observe increased performance when I use my PC, then a tweak is not of much value.

    Perceptions can be tricky. If you believe an action will result in a concrete performance gain, then it may appear to do so. It is called a placebo effect.
    Deadeye81

    "We make a living by what we get, we make a life by what we give." Sir Winston Churchill

  6. #5
    Bronze Lounger DrWho's Avatar
    Join Date
    Dec 2009
    Location
    Central Florida
    Posts
    1,501
    Thanks
    30
    Thanked 205 Times in 163 Posts
    I've used that tweak on countless # of PC's and it works.
    I'm using it right now, on my own XP PC which is several years old and runs
    faster than a lot of the new computers I have to install today.
    By the way, it also works on Vista and Win-7.

    But if you don't have ample ram, don't even think about it.
    That's about 2 gig's for XP and 3+ for Vista and Win-7.

    Two points to consider as to why it might just work:

    One, the OS has to access the Kernel sometimes many times a second.
    If it's on the hard drive, that access takes a LOT longer than if the Kernel is in FAST ram.

    Two, the ability to use that tweak didn't just fall out of the sky. It actually was programmed
    into Windows by the Microsoft Programmers.
    Having the kernel on the hard drive is just another one of those SAFE DEFAULTS that MS puts
    into Windows to allow it to operate on just about any PC, even the ones with very little RAM.

    I don't care what you try to do to improve your PC, someone will come along and tell
    you you're crazy. I just don't pay any attention to that stuff any more.

    I just do what I do and don't pay any attention to the Nay-Sayers.

    I regularly get comments from my customers, like, "WOW! It didn't run that fast when it was new!"
    So I know I'm doing something right.

    Cheers mate!
    Last edited by DrWho; 2011-10-02 at 18:27.
    Experience is truly the best teacher.

    Backup! Backup! Backup! GHOST Rocks!

  7. The Following User Says Thank You to DrWho For This Useful Post:

    RangerPHX (2013-05-29)

  8. #6
    Administrator
    Join Date
    Mar 2001
    Location
    St Louis, Missouri, USA
    Posts
    23,584
    Thanks
    5
    Thanked 1,059 Times in 928 Posts
    The only real way to tell is to run performance tests. Check out any reputable hardware site such as Tom's hardware for the tests they run when comparing systems.

    This is a much discussed subject which often generates many opposing posts. I've never seen any supporting performance documentation from those who support the tweaks. All I ever see is someone (or several people) say "It never ran that fast before". That is not proof to me. I strongly agree with Deadeye81's comment about perceptions.

    Joe

  9. #7
    4 Star Lounger
    Join Date
    Jun 2010
    Location
    Littleton, Colorado
    Posts
    492
    Thanks
    33
    Thanked 6 Times in 6 Posts
    Dr. Who, it was your similar comments in another thread that lead me to try these tweaks. While I find your logic to be reasonable, I haven't as yet noticed a performance difference. That's not to say that there isn't an improvement, but until I can actually cause the computer to frequently access that part of the kernel that was previously kept on disk instead of RAM, I can't really say that it's making a difference.
    But I'd certainly like to get some empirical data. Whether or not it's noticeable, if there is a consistently measurable performance improvement, no matter how small, I would like to know.

    Toward that end, I have the latest version of Passmark's Performance Test software. However, I don't know if any of the tests that it can run would actually exercise the system sufficiently to give me a definitive answer to this particular question.

    This leads to another question (besides my original one, which has not been answered yet); is Performance Test the right software to resolve this issue? I will also check out Tom's Hardware site, as Joe suggested, but unless it's a freebie, I'm not inclined to spend any additional money of testing software just to resolve this one point.

    Thank you all for joining this dialogue.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •