Vista services and optimizations for oqo 02

[Guest]

[vting]

Of course why didn't I think of it.

Now I just have to find them and delete them

Thank You!

[Opus]

found this so far, going to try it myself...

http://windowshelp.microsoft.com/Windows/en-US/Help/1264bc24-72a8-48aa-84e3-a355327139d91033.mspx

[vting]

Thank you for finding a solution.

That I think is good enough unless there is a third party program that can do batter.

Thanks for helping out.

[PlacidoDomenech]

I am actually using ProcessTamer free utility and vista runs fine.

[stuartguthrie]

   - Virtual memory - Change - Custom Size (don't check "automatically manage paging file size for all drives") - initial size/1536 - maximum size 1536 - set

Hi all,

I was wondering if someone could please explain what this does? I noticed that mine was set to 'Automatically manage paging file size for all drives' so I've changed it to Dennis' suggestion of 1536 / 1536.

Thanks in advance.
Stu

[Dave P]

If you allow Windows (XP as well as Vista) to manage the page area, it will attempt to optimize space on the drive by removing pages as they "decay" through lack of use. When more pages are active than it has space for, it will attempt to increase the overall page area size.

When you fix the page area size, Windows reverts to filling up the page area and then, as new pages need to be added, it will remove the oldest pages to make room for them. This results in Windows having to do less tracking, fewer reads and writes, and less "thinking".

This only works if you set the size near or over the recommended maximum. It also results in that space being taken out of the available space pool so that other programs cannot use it. Now that we speak in terms of GB rather than MB, the latter tends not to be an issue.

[stuartguthrie]

If you allow Windows (XP as well as Vista) to manage the page area, it will attempt to optimize space on the drive by removing pages as they "decay" through lack of use. When more pages are active than it has space for, it will attempt to increase the overall page area size.

When you fix the page area size, Windows reverts to filling up the page area and then, as new pages need to be added, it will remove the oldest pages to make room for them. This results in Windows having to do less tracking, fewer reads and writes, and less "thinking".

This only works if you set the size near or over the recommended maximum. It also results in that space being taken out of the available space pool so that other programs cannot use it. Now that we speak in terms of GB rather than MB, the latter tends not to be an issue.

Hey Dave,

Thanks for the reply. Please excuse me but I'm still a bit confused. 

So, is it a good thing to manually manage the paging file size? If yes, is 1536 / 1536 what I want it set to? How do you have yours configured?

Thanks again,
Stu

[Dave P]

So, is it a good thing to manually manage the paging file size? If yes, is 1536 / 1536 what I want it set to? How do you have yours configured?

Yes, it is generally a good thing (as far as system performance) to manually set the paging file size. The downside is that you loose some potential disk space but this is not usually an issue with gigabytes of storage.

For myself, I used what Windows recommended as the maximum at the time that I set it which was 1429MB. Your proposed setting of 1536MB sounds fine. I have heard 1.5 times RAM as a recommendation but I don't know if that has any basis in fact. That, I assume, is where the 1536MB came from. It's 1024MB (which is 1GB) times 1.5.

[stuartguthrie]

Hey Dave,

Thanks for that info. Sorry it's taken me so long to reply (I've been bogged down with studying for exams so I haven't been on OQO Talk much).

Thanks again,
Stu

[Ronc]

Your proposed setting of 1536MB sounds fine. I have heard 1.5 times RAM as a recommendation but I don't know if that has any basis in fact. That, I assume, is where the 1536MB came from. It's 1024MB (which is 1GB) times 1.5.

When I overrode automatic settings and used 1536MB there was a problem with tab delays in Vista/IE7.  I discovered it was caused by a couple of Java add-ons.  I could either disable them or add a Java control panel setting: Use the Java tab and view the "Java applet runtime settings"; fill in the Java Runtime Parameters box with "Xmx768". Apparently this sets the memory for Java at 3/4 RAM which is necessary when Windows can't adjust the virtual memory. (Discovered on Java troubleshooting web page.)

[Dave P]

Your proposed setting of 1536MB sounds fine. I have heard 1.5 times RAM as a recommendation but I don't know if that has any basis in fact. That, I assume, is where the 1536MB came from. It's 1024MB (which is 1GB) times 1.5.

When I overrode automatic settings and used 1536MB there was a problem with tab delays in Vista/IE7.  I discovered it was caused by a couple of Java add-ons.  I could either disable them or add a Java control panel setting: Use the Java tab and view the "Java applet runtime settings"; fill in the Java Runtime Parameters box with "Xmx768". Apparently this sets the memory for Java at 3/4 RAM which is necessary when Windows can't adjust the virtual memory. (Discovered on Java troubleshooting web page.)

I hadn't run across that but it sounds reasonable. XMX sets the maximum memory for the Java Heap. Just one thing, though, while I would guess that most implementations would recognize "Xmx768", as I recall the strictly proper format would be "-Xmx768m" (without the quotes).

[benjitao]

The problem might not be the operating system at all.  The C7 chip is a In Order CPU.  These are great for a car, but they do not work well in a multi-processor environment like with a GPU, or even a southbridge controller chipset.  They need their processes to happen sequentially, and sometimes they send instructions to the southbridge or the GPU that need a follow up from the CPU, but the CPU is waiting for the GPU or southbridge to send it back what it wants, but the GPU is waiting for the CPU, so it locks up.  In Out of Order no sequential CPU's the instruction that the GPU is waiting for will be handled by the CPU and it will get what it needs to finish a former instruction.  Modern operating systems are coded dependent on Out of Order cpu's.  The drivers for many video cards are coded depended on out of order CPU's as well.  You have to patch the operating system and the drivers to make everything work nicely with the dumb in order CPU.  I bet an analysis would reveal a number of places where sloppy code takes advantage of Out of Order processing rather than keeping better track of its processes.  Unless you are going to taylor the operating system for an In Order processor then you are going to have problems with it.  Linux might have a better codded kernel, but it is not perfect anymore either, specially with a GPU.  Unless you manage your code better and how the GPU, memory controller, southbridge controller, CPU communicate you will never gt the OQO to run as well as an out of order CPU would, because they just are not designed to work in a multi-processor environment like your modern computer with its GPU and controller chips.  They are better as computational slaves that as CPU's themselves.  The Atom Oqo's are not likely to be much better.  OQO's best bet and Intel's for that matter if they want to use in order processing is to get a team to make a tight Linux kernel.  Windows is too too out of order dependent.

[kitenski]

Evening all,

Is there a similiar XP services and optimization guide for XP anywhere?

regards,

greg