Which OS for 3DMark06?

[vega22]

i dual boot xp and vista 64bit and xp is miles faster, for everything. having said that i do like vista its just not as fast as xp but it is a nicer looking desktop with some real good features that xp just cant do.

i lose around 500 in 3dmark06 but in games its alot more. in cod4 i can get about 150/160 fps in vista but with xp i will get 300+ crysis i will get 60/70 fps on high but switch back to xp and we are talking 100+.

i think part of this is down to the dx9 emulation of vista/dx10.

 

[Graogrim]

XP was designed around never having any spare memory and to be mean with it.. he he

I wouldn't go THAT far.

Seriously, have you tried adjusting your computer's memory usage to favor the system cache over applications? If you've got gobs of RAM to spare the more aggressive caching might prove beneficial. XP's system cache works a lot like SuperFetch in Vista, with the main difference being that it's not proactive. If you load a large app in XP, close it, and then reopen it, you'll notice that it restarts very quickly. That's your system cache in action. Another really neat thing is that once program code is in the system cache, it doesn't have to be moved to another part of RAM to run it--it can actually be executed directly from the cache. One of the more powerful implications of this is that you can have multiple instances of a program running simultaneously, and though they'll each need to allocate their own workspace they all run from the same place, thus avoiding redundancy.

 

[ManofGod]

i dual boot xp and vista 64bit and xp is miles faster, for everything. having said that i do like vista its just not as fast as xp but it is a nicer looking desktop with some real good features that xp just cant do.

i lose around 500 in 3dmark06 but in games its alot more. in cod4 i can get about 150/160 fps in vista but with xp i will get 300+ crysis i will get 60/70 fps on high but switch back to xp and we are talking 100+.

i think part of this is down to the dx9 emulation of vista/dx10.

What resolution are you running those games at, 1024 x 768 or less?

Joe

 

[trog100]

I wouldn't go THAT far.

Seriously, have you tried adjusting your computer's memory usage to favor the system cache over applications? If you've got gobs of RAM to spare the more aggressive caching might prove beneficial. XP's system cache works a lot like SuperFetch in Vista, with the main difference being that it's not proactive. If you load a large app in XP, close it, and then reopen it, you'll notice that it restarts very quickly. That's your system cache in action. Another really neat thing is that once program code is in the system cache, it doesn't have to be moved to another part of RAM to run it--it can actually be executed directly from the cache. One of the more powerful implications of this is that you can have multiple instances of a program running simultaneously, and though they'll each need to allocate their own workspace they all run from the same place, thus avoiding redundancy.

well yes after years of doing the opposite i did just that.. what it dosnt do is keep it there.. i have given the cache priority over programes.. it dosnt make a deal of difference.. most of my 4 gigs sits doing nothing.. just like most of my 2 gigs used to do..

this is my normality.. and thats after just shutting down a memory hungry game and the system as been up for twelve hours.. ..

after a few hours running i would like to see at least a gig of my memory used.. i dont and still have to wait for oft used apps to load from the hardrive.. XP is tooo mean with spare memory.. it clears its cache way to quickly and when it dosnt have too.. this used to be good now it isnt.. it slows things down..

trog

ps.. and just for a laff i have tried loading five games at once into memory and playing the last one.. i really do have lots of spare memory.. i just wish XP made better used of it.. he he

 

[trog100]

the power of XP.. only joking..

four biggy games loaded at once..

fear..

crysis..

gears of war..

conflict denied opps..

the one running is conflict denied opps showing 58 fsp.. the other three are alt/tabbed into the background.. 3.3 gigs of ram in use..

the cpu usage is whatever is needed to keep all four games running in the background.. any more and it would run at of ram and have to start swap filing..

just to make the point as regards what i have sat there doing nothing most of the time.. it might make the other odd point.. buggered if i know what thow.. he he he

trog

ps.. the four games are still sat there and i can browse about and not notice.. praps it proves i have way more computing power and memory than i need.. dunno..

 

[Graogrim]

In your first screenshot I see you have a total of about 3.5 gigs of RAM available under XP. I see about 2.8 gigs listed as available, and about 2.5 used for the system cache. The fact that those numbers don't add up is significant. It's important to note that memory that's "available" isn't necessarily going unused. It's just that it can be freed on the fly to provide memory to programs.

So don't sweat it. Right there you've got most of your "available" RAM holding information for the cache. It's all good.

 

[trog100]

window uses virtual memory which is a mix of ram and hardive.. for some silly reason the more ram u have the bigger the swopfile is that gets created..

XP seems to decide on one to one.. so with 4 gigs of ram it sets aside 4 gigs of hardrive space.. it then refers to this as available memory.. 8 gigs..

my swopfile now called pagefile.sys is set to default at 2 gig.. when task manager say memory available it includes this 2 gig hardrive file..

techically windows 64 bit has virtual memory limited by the hardive space.. sadly the hardrive part of its virtual memory is that slow its no longer really valid..

now if it was done sensibly.. it would use its spare ram instead of its prefetch.. simply leave (or preload) oft used apps in memory ready for instant use.. waiting for the hardrive should rarely happen..

i think one of the vista advantages.. praps the only real one.. is that it does use its spare memory (ram) to do this.. XP just just leaves spare ram standing idle.. which is my only gripe with it..

hence all my blatherings.. he he

trog

 

[Graogrim]

window uses virtual memory which is a mix of ram and hardive

The information on your pagefile usage is reflected in the "commit charge" section, which shows your total allocated memory, both physical and paged.

.. for some silly reason the more ram u have the bigger the swopfile is that gets created..

XP seems to decide on one to one.. so with 4 gigs of ram it sets aside 4 gigs of hardrive space.. it then refers to this as available memory.. 8 gigs..

You can easily control the behavior of your swapfile, and set whatever limits on it that you please, both minimum and maximum. You can even disable it completely if you so wish, though I don't really recommend it.

my swopfile now called pagefile.sys is set to default at 2 gig.. when task manager say memory available it includes this 2 gig hardrive file

Not exactly. I'm referring to the "Physical Memory" section, which doesn't include pagefile usage. To get the complete picture you have to look at both the Physical Memory section and at your Commit Charge, which reflects virtual memory (your physical memory combined with your pagefile(s)). To help with the picture I'll use the current numbers from my own Task Manager:

Commit Charge (K)
Total: 973344 (this is the current total of allocated virtual memory(pagefile+physical))
Limit: 4034876 (this is the total amount of virtual memory (pagefile+physical))
Peak: 994400 (this is the highest amount of virtual memory that has been allocated this session)

Physical Memory (K)
Total: 2096492 (total usable physical RAM. pretty straightforward)
Available: 1024272 (physical RAM currently available, can include some of the space in the system cache, this is not necessarily fallow memory)
System Cache: 407488 (this is the amount of physical RAM currently used by the system cache)

I have two gigabytes of RAM installed, this is reflected in the total under Physical Memory. My pagefile is currently configured to occupy between a minimum of 2046 MB and a maximum 4092 MB. At the time these numbers were polled it was sitting at the bottom end of that range. The sum of my pagefile and my physical memory (less some nonpageable areas) is shown under Limit in the Commit Charge section. Since the pagefile can grow up to nearly four gigabytes if needed, the Limit is subject to change.

simply leave (or preload) oft used apps in memory ready for instant use.. waiting for the hardrive should rarely happen..

This describes Vista's SuperFetch feature. But here's the catch. TANSTAAFL (there ain't no such thing as a free lunch). It's true in thermodynamics and it's definitely true in computing. One way or another, you WILL be waiting for the hard drive to load that data. It's just a matter of timing. A SuperFetch enabled system will have more disk activity at startup, and that's where the loading cost is paid. You can interrupt that process of course, but then the benefit offered by SuperFetch is offset.

XP just just leaves spare ram standing idle.. which is my only gripe with it..

Except that it doesn't really. Although XP doesn't proactively load the cache (except for standard read-ahead), it won't purge the cache needlessly, either.

So relax. You may not have SuperFetch, but your system cache is doing you some good.

 

[trog100]

i aint gonna argue dude.. bin there done that before concerning windows and its memory use.. u might think XP makes good use of spare ram.. i dont think it does.. i think its best we leave it at that..

i do agree about the free lunch but only to a point.. one load into memory is essential.. repeated loads isnt and to a large extent could be avoided with more efficient memory management.. XP just dosnt have it.. its outdated priority is based on conserving it..

Vista and its poor sales has done us one big favour thow.. plenty of cheap DDR 2 for XP even if it dont make good use of it.. he he

trog

 

[Graogrim]

It's cool, I'm not arguing either. But you have just tried to educate me with at least two factual inaccuracies.

I take a lot of pleasure in learning about the technical side of computing, and helping other people learn. I might throw out walls of information, but I only do it because I enjoy the discussion. If you really don't want to talk about it though, I won't push. I can only urge you to research the topic further on your own. Heck, DON'T take my word for it. I'm just some random schmuck on a forum. Read up from a real authority like Microsoft's own technical references.

 

[trog100]

It's cool, I'm not arguing either. But you have just tried to educate me with at least two factual inaccuracies.

I take a lot of pleasure in learning about the technical side of computing, and helping other people learn. I might throw out walls of information, but I only do it because I enjoy the discussion. If you really don't want to talk about it though, I won't push. I can only urge you to research the topic further on your own. Heck, DON'T take my word for it. I'm just some random schmuck on a forum. Read up from a real authority like Microsoft's own technical references.

i have researched the subject extensively over the years.. most of whats written is years out of date and the fact the results of my conclusions and personal tests and observations happen to differ from yours proves nothing..

why on earth do u think i would take anybodies word for it never mind yours.. he he

plus when any guy starts debating with the use of "selective" quotes i stop..

the last time i did this memory and how windows uses it debate stuff was with a dude called alec star.. or something like that..

it went on for page after page.. its no doubt still there in the archives..

he claimed windows had to have a pagefile.. i disabled mine to prove otherwise.. he claimed windows created another one.. i said where.. he said it was a"hidden" one.. i said hmmm..

the real problem i had with this particular debate was a simple one.. he had the jargon and the references to back up his argument.. i simply had what i saw or could reproduce.. ..

he quoted "experts".. google around and its all there.. it can all be pointed at..

all i have is simple practical tests i conduct myself..

i believe in my simple practical tests.. if i cant reproduce what the "experts" say should happen i doubt the "experts"..

when half a dozen "experts" disagree i dont wonder which one is right.. i assume none of them are..

my name is thomas and i am a doubter..

live well and prosper..

trog

ps.. mind u i am curious about the two factual errors.. i thought we had more than that.. he he he

 

[Graogrim]

I'm sorry if you take exception to my "point for point" posting style. I'll try to remember that in any future dealings with you. Personally, I find that it makes for a clearer exchange of ideas, but if you find it threatening then I'll avoid it.

At this point I'm inclined to say that you were right. We're done. When one party of a discussion becomes convinced that an argument is taking place, it becomes a sort of self-fulfilling conclusion. You've as much as told me that nothing I say and no references I provide will help, and you're unwilling to go looking for more on your own. That's religion, and there's no getting past that.

My name is Matthew. I'm not a doubter, but I am a skeptic. I'm pleased to meet you Thomas. I'm sorry we couldn't make a profitable discussion of this.

 

[Laurijan]

Try to stick to the topic - whats faster in 3DMark06.
My tests says XP 32Bit is faster then XP 64Bit or Vista - XP 32Bit has a about 400pt faster scores in 3DMark then both later ones.

 

[trog100]

Try to stick to the topic - whats faster in 3DMark06.
My tests says XP 32Bit is faster then XP 64Bit or Vista - XP 32Bit has a about 400pt faster scores in 3DMark then both later ones.

somewhere back in the thread most folks have agreed with your findings..

XP 32 bit is faster in 3dmarks..

trog

 

[FatForester]

trog100, for the love of everything holy please use complete sentences. I literally stopped reading the thread because of it.

 

[trog100]

trog100, for the love of everything holy please use complete sentences. I literally stopped reading the thread because of it.

my habits are far too ingrained for me to change em..

But out of curiosity, do you stop reading because you fail to understand the things I write; or simply cos u hate bad grammar and forum speak.. ????

trog

 

[warhammer]

I thought this may be of intrest on the VISTA and XP argument.

Another take on Vista vs. XP benchmarks

My experience is the same. In fact, it appears that Vista’s designers have made a conscious choice to favor smooth, consistent performance over raw speed. The latter makes for more satisfying benchmarks, but it can also result in annoying performance glitches in day-to-day use.

http://blogs.zdnet.com/Bott/?p=369

 

[asb2106]

I thought this may be of intrest on the VISTA and XP argument.
http://blogs.zdnet.com/Bott/?p=369

Is file copying really a critical performance benchmark? If it takes me 10 seconds more or less to copy a group of files, I truly don’t care. For mainstream business use, there’s no practical difference between a job that takes 5:52 and one that takes 6:18, especially when the copy operation takes place in the background while I busy myself with other work.

I really agree with this line here,

There are more background operations that use disk resources, but in a real life senario what difference does 10 seconds really make. I could care less! And as was mentioned in the article, vista has gone for more of a streamlined approach instead of raw speed, and I can appreciate that.