Printable View
Apologies if this has been covered before, but I don't see it anywhere. I was looking for performance tweaks for a game when I came across this. I don't have a dual core CPU (yet, but I may soon), but these settings look like they could make a big difference to those that do have them.
linky
The gains are quite impressive...
Of course I'm using a P3 1ghz at the moment, so not much chance of it helping me...
Nice find!
I wonder why microsoft wouldn't want this pushed out to everyone. The only thing I can see is that it may decrease battery life.
http://support.microsoft.com/kb/896256
Lots of info in there for those who want to read it.
I would say you would be foolish not install the fix if you have a dual core or hyperthreaded processor.(edit)
Great source, Xiahou. Though I'm unable to test it, I really appreciate such treasurely findings. :bow:
I installed this before, but didn't do any before and after tests.
There was also another patch from Microsoft concerning Dual Core Performance and energy saving mode. IIRC I didn't do any registry editing on my current install, just applied several patches or so.
I might want to check that again once I get back to my PC.
Could be a placebo effect, but my PC seems faster. :wink3:
Thanks for the tip, X-man.
Hmmm, interesting. BlackAxe's link too. Thanks, I'll have a good read with these :2thumbsup:Quote:
Originally Posted by Xiahou
-edit-
I've hardly seen a difference, with the C2D running at 2.13Ghz, 1066 FSB, etc. At least I haven't in games. Now, a single-core CPU with higher clock speed would outperform a dual-core in games, unless they optimize them for it.
Does this work on core 2 duos or are they a horse of a different color?
I'm not entirely sure about this because most single cores rely on outdated technology and the Core 2 Duos have outperformed the fastest AMD single cores in reviews I've seen even if the game used only one core. And a dual core will also give better performance with other programs running in the background, because it can "outsource" all the small background stuff to the single Core and concentrate on the big game with the other core.Quote:
Originally Posted by Bijo
Hmmmm. I'm not entirely sure either. If these games only use one core, which is usually the case currently, and the dual-core CPU uses one for the other background tasks, and it is supposedly faster than a single-core with higher clock speed, then I wonder what reviews you've seen. Do you have links, or better yet: is there any solid hard proof?Quote:
Originally Posted by Husar
Hmmm, as far as I know, it seems logical to assume that if one 2.0Ghz core runs a game, and the other 2.0Ghz core runs other stuff, a single-core CPU of 3.0Ghz should beat the dual-core's game performance, yet something tells me it doesn't necessarily :dizzy2: But fact is: still one core is being used for the game, meaning the superfast single core "should outperform" the used core of the dual core.
And if that's the current purpose of dual cores with games, to still run it with a slower single core, while tasking stuff with the other, that's in my book kinda messed up. Those games should use both cores effectively, and Micro$oft should stop promoting software development and enforcing "consumer needs" that require the system to run unnecessary stuff in the background. Avoid bloated software with bloated code.[/rant] :sweatdrop:
The thing about other software was meant in a positive way because I like to keep Fraps, ICQ and MSN open in the background, sometimes even Opera.
Concerning performance, the fastest single core was the FX-57 IIRC(couldn't even find that in a shop anymore) and it was outclassed by the FX-60 in almost all gaming tests I've seen(don't know whether all the games were dual core optimized) and the FX-60 is seriously outclassed by the Core 2 Duos and that is why there is a lot of reason to think that a single core of a Core 2 Duo will outclass any available single core, single cores are all based on outdated technology anyway, there is nore Core 2 Single and i think the new K8L platform from AMD will be all dual and quad cores, so the era of single cores is definitely over anyway. And games are starting to use the multi cores anyway.
Ive been told to ignore the ghz rating in comparing single core to dual, especially when it concerns Core 2 Duos. Even at a lower rating they say it runs faster and cooler than a single core witha much higher rating.
I know it's not completely authoritative, but I like the CPU charts at Tom's Hardware- they make it easy to compare a range of processors under various tasks and get a general idea where they stand. Here's the link for the CPU charts if anyone's interested.
I'm guessing it would be the same for any multi-core processor. :shrug:Quote:
Does this work on core 2 duos or are they a horse of a different color?
Notiice that in general the Core 2 Duos out perform all the other proccesors no matter what the Ghz rating. My E6400 is only rated at like 2.18 yet out preforms those with a ratting of 3 or even 4 ghz. It seems to me this is a whole new ball game and as I said you cant just go by the ghz rating when comparing single core of even 1st generation dual cores to the Core 2 Duos. But then I dont know much about this stuff and am going by what i see.Quote:
Here's the link for the CPU charts if anyone's interested.
If they're using a whole new chip architecture for the new dual core processors then it will be exactly like trying to compare an AMD processor with a Sledgehammer/Winchester or what have you core that runs at 2.2ghz and a P4 that runs at 3.2ghz. The AMD in this case will give you better performance iirc. This particular example may not be entirely correct because it's been several months since I've seen my lovely 3500+ and I don't remember exactly if it outclassed a P4 3.2, but it outperformed a 2.8 quite handily in any case.
Anyway the point is the new chip architecture will quite possibly totally change things and the new chips probably run more efficiently than the old ones allowing them to run at a lower frequency and produce the same or better results.
I probably should have stayed out of this since I've been away from my baby for so long, but I'm sure you'll all forgive me and my point is still hopefully valid despite my outdated information.
I took out some words with [...] that I thought unnecessary to our argument :)
But I've read that in single-threaded programs (which is what many games were and are still using) the FX-57 would win.Quote:
Originally Posted by Husar
Hmmmm, I find this peculiar logic in itself. Because the C2D (assuming the information is correct) would beat the FX-60, doesn't necessarily mean one C2D core would beat a single-core CPU. It still depends on whether the software is optimized, which is an important factor.Quote:
and the FX-60 is seriously outclassed by the Core 2 Duos and that is why there is a lot of reason to think that a single core of a Core 2 Duo will outclass any available single core,
They are indeed starting to use more cores, true, now, implying they didn't do so much before. Single cores use outdated technology, so did/do the games; it didn't/doesn't matter. Dual cores use new technology, but if the game doesn't use this technology effectively, what's the point (besides future-proofing your system)?Quote:
single cores are all based on outdated technology anyway, [...] And games are starting to use the multi cores anyway.
Still, fact remains: a slower core of a dual-core CPU is used for a game, or a faster single-core CPU is used for a game. When I look at those "recommended specs" I usually see something in the sense of Pentium 4 3.0Ghz or equivalent (or something like that, or higher speed). I never see anything on the game box or in the manual or readme files about dual cores, which are supposedly better CPUs, because those games aren't designed for it - they still use one core.
It then makes sense to think that a fast 3.xGhz single-core CPU would beat a 1.xx/2.0Ghz single core of a dual-core CPU.
I think we should find some kind of list of games that are optimized for dual-core CPUs :yes:
I also think we should take into consideration other factors of the system, especially hard disks. For instance: if you have dual-core, run a game with background tasks, the hard disk plays an important part.
-edit-
About the dual-core tweak:
I've used System Restore to have my system back to how it was before I used the "tweak" and it now seems faster. When I actually did the tweak back then it really seemed slower.
-edit2-
http://www.hardforum.com/showthread.php?t=983781
Found another link to this tweaking thing.
Let me have a look...Quote:
Originally Posted by Bijo
Well, picture from the Core 2 Duo test from www.k-hardware.de
http://pics.krawall.de/4/core2_x6800_3dmark03_cpu.gif
Now this is 3DMark 03 because it uses only one core for benchmarking, the Core 2 Duo E6400 and lower are probably slower than the FX-57 when using only one core but I almost never think of them when talking about the Core2Duo.:sweatdrop:
What I meant was that a single core of an FX-60 would beat an FX-57, maybe that was a wrong assumption because the FX-62 is actually the same(as in GHz per Core) as the FX-57 as can be seen in the graphic above. Well, it still outperforms the FX-57 because background stuff can be outsourced to the second core.Quote:
Originally Posted by Bijo
So a single core of an FX-60 may not outperform an FX-57, but the Core 2 Duo from E6600 and up definitely does.
Erm, wrong, the new technology does not need to be used by the game because it simply enables the processor to calculate simple tasks faster. The new technology is used by the hardware itself, not by the software. For example the Core 2 Duo can calculate more basic operations per Hertz than the AMDs can. Here they posted some Intel graphics which are supposed to explain that, but don't ask me, I'm not an expert with that stuff either.~;)Quote:
Originally Posted by Bijo
There are also things like more cache, faster bandwidths here and there etc.
No, as shown above this is not the case.Quote:
Originally Posted by Bijo
The recommended specs are basically for noobs anyway and most experienced users know that they can't be trusted in any way.
A 4 GHz CPU would be one of those really old Intels they sell at bargain prices now because they are so bad, get so hot and are even slower than the FX-57 we talked about before. Just look at the Pentium 4 670 in the graphic above...
Well, that's not a bad idea, I often check task manager to see what exactly they use, so far I know:Quote:
Originally Posted by Bijo
Gothic 3 - definitely both cores to full potential
Medieval 2 - one core fully used, the other only partly
Supreme Commander - the demo and beta used only one core
Well, not much, maybe I should check more in the future.:sweatdrop:
Well, that always depends a lot on the game and the background task. For example I often play Medieval 2 or Mount&Blade while downloading since none of these uses the HDD to any great extent. Gothic 3 for example is a very different matter, I try not to run anything in the background because it basically eats your RAM and HDD performance by constantly streaming the world(that's also what the second core is supposedly used for) I'm not sure about other games, but there are many factors playing into this and with many games, a lot of RAM takes a lot of work away from the HDD(except of course, the streaming of huge maps/locations/worlds)Quote:
Originally Posted by Bijo
I got rid of that problem more or less by migrating to Vista now.Quote:
Originally Posted by Bijo
Want to install XP as a secondary, but my CD is packed with Service Pack 0 and I lack a Floppy and burner and have an SATA HDD now so i have to find someone to create me a nice XP SP2 installation CD. And while we're at it, could this tweak here also be streamed onto a new XP CD?:inquisitive:
I just hate installing all the patches and fixes for XP all the time, from creating partitions to installing and running the first programs took me less than an hour with Vista and it still runs fine so far.:2thumbsup:
I was actually thinking more of that kinda stuff :P Two cores with lower speeds, versus one core with higher speed. The way you said it in the previous posts was like you included every C2D (or general dual-core CPU) in your argument. And the C2 X is of course a step higher over E6600 and up, which are higher than E6400 and below.Quote:
Originally Posted by Husar
Still, when I read about games here and there, there are mixed reactions. Sometimes they see an improvement, sometimes they don't. I wouldn't say the games nowadays have such simple tasks, though. Your point sounds good, but I can't help but to say that if a game is not optimized for dual-core use.... etc. That picture, hmmmm.Quote:
Erm, wrong, the new technology does not need to be used by the game because it simply enables the processor to calculate simple tasks faster. The new technology is used by the hardware itself, not by the software. For example the Core 2 Duo can calculate more basic operations per Hertz than the AMDs can. Here they posted some Intel graphics which are supposed to explain that, but don't ask me, I'm not an expert with that stuff either.~;)
There are also things like more cache, faster bandwidths here and there etc.
In any case, indeed, those other factors must be considered.
Btw, that link is in German! I can't read German that well and quickly :laugh4:
But here I've mainly said that on game boxes/manuals/readme's usually there's nothing to read about dual-core optimization, probably because there is no TRUE dual-core optimization. What does "No, as shown above this is not the case." exactly imply? Do you mean the picture? If so, it only shows performances of different processors in single-core mode - not whether people use certain processors.Quote:
No, as shown above this is not the case.
The recommended specs are basically for noobs anyway and most experienced users know that they can't be trusted in any way.
A 4 GHz CPU would be one of those really old Intels they sell at bargain prices now because they are so bad, get so hot and are even slower than the FX-57 we talked about before. Just look at the Pentium 4 670 in the graphic above...
:dizzy2: *DROPS DOWN, DEAD!*
I must ask you the following question: did you FABRICATE that picture?! LOL :laugh4:
Hmm, doesn't Vista still cope with performance problems?Quote:
I just hate installing all the patches and fixes for XP all the time, from creating partitions to installing and running the first programs took me less than an hour with Vista and it still runs fine so far.
____________________
-edit-
ABOUT THE ORIGINAL POST REGARDING DUAL-CORE PERFORMANCE TWEAKS:
http://support.microsoft.com/kb/895980
That link sounds fishy to me (even though it doesn't make a sound) or it smells fishy to me (even though it has no smell) :P In any case, there are mixed reactions as to whether the thing works. Seems that AMD Optimizer adds /usepmtimer to BOOT.INI if I'm correct. I'm skeptical and don't know at all what to do now with this thing.
I also installed the latest available BIOS long beforehand that supposedly would recognize and properly support "newer processors".
Yes, but the E6600 still has 400MHZ less per core, yet outperforms the FX-57.Quote:
Originally Posted by Bijo
Never meant to say that the AMD dual cores are faster than their single core pendants, sorry if I made that impression.
Well, whether a game is "optimzed" or not is more or less a matter of whether it is multithreaded or not. Gothic 3 for example has seperate threads for sound, streaming, the game itself, probably physics etc.Quote:
Originally Posted by Bijo
Personally, I always wondered how it was ever possible to program a single threaded game, but I'm just a noob at programming(I'd have created a new thread for each soldier most likely~:rolleyes: ).
Oh, I posted it just for the pics, because I didn't want to put them into my post. the descriptions on the pics should be in english.Quote:
Originally Posted by Bijo
Gothic 3 has no such message on the package, but dual core optimization was actually mentioned quite early in the development and the community manager once said that without a dual core, there will always be reload lag because the single core will also have to do the streaming of the world. I cannot really tell how true that is because I never ran it on a single core.Quote:
Originally Posted by Bijo
Yes, of course.:dizzy2: :2thumbsup:Quote:
Originally Posted by Bijo
It copes very well, haven't noticed any significant difference so far but didn't calculate any averages or some such thing either.Quote:
Originally Posted by Bijo
I'm not 100% sure about this, but it's my understanding that even if a game isn't optimized for dual core processors, you still benefit as your computer can multitask the game with other things (even just running the OS in the background) better.
Plus there is next to no reason I can see for not getting one of the C2Ds if you are looking for an upgrade.
That there in bold would roughly have prevented most of our earlier arguments from even taking place if stated earlier :P That and what you said about the E6400 and lower "probably" wouldn't beat the FX-57 using one core.Quote:
Originally Posted by Husar
That picture, though, is quite incomplete. For the sake of completion I suggest another complete picture to replace it, and maybe even another benchmark like 3DMark05 and/or SuperPI.
I knew it!!! :smash:Quote:
Yes, of course.:dizzy2: :2thumbsup:
But the argument regarded more games in general, not only one or one called Gothic 3.Quote:
Gothic 3 has no such message on the package, but dual core optimization was actually mentioned quite early in the development and the community manager once said that without a dual core, there will always be reload lag because the single core will also have to do the streaming of the world. I cannot really tell how true that is because I never ran it on a single core.
Still fact is, a fast single-core CPU on a good fast system will play many games well. And games are the best benchmarks, and many games aren't dual-core optimized (let alone single-core optimized, btw :P). But you know what? We both ain't no specialists or experts, and the proof you present isn't hard enough if you ask me. Let's get this down with once and for all! Can somebody who knows some real **** step in and provide us with solid hard proof regarding single-core games and CPUs?
I'm not pleased yet, but die Diskussion war gut genug :bow:
Excuse me, but what is Intel Dual Core?A kind of super-microprocessor?
If yes, they stole a part of an oldie idea i had.
Its 2 proccessors in one. People, Ive been told that one of the main reasons they blow away the competition is they are the only cpus that run at 1066. Have any of you seen the new ones in the micro towers? Incredibly small.Quote:
Excuse me, but what is Intel Dual Core?A kind of super-microprocessor?
Hmm, got to be honest, not quite sure I've understood this thread. Now my question is, will this thing increase the performance of my CPU?
I have a "mobile dualcore intel core duo T2400"...
Thanks for any help :)
May benefit, but will not harm.Quote:
Originally Posted by HoreTore
Well, I tried this on my new PC and am disappointed to say that it made little difference. :shame:
I saw an improvement of about 30 marks in 3dMark06... which is pretty insignificant. I wonder if installing the AMD CPU driver makes for most of the performance gains? I did that ahead of time, since it was included on my mobo CD- that could possibly explain why I saw so little difference. :shrug:
That is why you didn't get any improvement. You already had it installed via AMD's driver/optimizer.Quote:
Originally Posted by http://forum.notebookreview.com/showthread.php?t=60416
I've performed these tweaks on my new E6300 PC & benchmark before & after with no discernible difference in performance. Also, during the CPU test on 3dmark06, my FPS hovered around 0-1 fps, surely this can't be right? Anyone with any clever ideas?
The 3dMark06 murdered my CPU as well. Regardless, the CPU marks were more than doubled on my new 4400+ x/2 than they were on my old 3600+ Venice core. The Venice ran the cpu test at a painful to watch 0 fps, whereas the new CPU at least kept it running between 0-1.Quote:
Originally Posted by HughTower
The difference in on screen FPS wasn't much, but it did run noticeably smoother- if still very slow, and the CPU score increase was quite significant. :shrug: