• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Micro stuttering - disable core parking

wow great tip, some games when launched I use to get the odd crackle and pop through the headset, not now and NFSHP feels way smoother.
 
Duff-Man said:
It would certainly be a very handy resource... I guess the "problem" is that listing all the possible incompatibilities could be seen as admitting that faults exist with their hardware/drivers, and so could be used as ammunition for 'the opposing side'. I guess some PR execs would take exception to this?

It's no excuse though really, and I'd always prefer to see a pragmatic approach to incompatibilities (which everyone knows are a fact of life in computing). Pretending that problems don't exist doesn't really help anyone.
Click to expand...

Exactly, that would be my main guess for it, though listing things like barely known MS multicore patches(it was definately different to the main one Vista had) is basically blaming MS so still ridiculous.

Theres really no proper excuse for all the companies to not help their customers better, and they really should all be refusing to play MS's WHQL game of dos and don'ts. Best for end user should be the only real goal here, not best for some arbitrary list of rules MS has imposed.

Oh well, I wait on for those two at least and maybe others, to buck the trend and make a list of good/bad patches/programs/helpful utilities to get the best out of their hardware. I think I'll be waiting for a LONG time for that to happen though.

On top of that, game dev's need a kick in the ass, the amount of games that crash as they hit 2gb memory usage on x64 systems as they haven't been set to address more than 2gb, and the amount of games that screen up FOV when its so incredibly simple for them to set a different default FOV for different aspect ratio's, everyones guilt of just being flat out lazy and not fixing easy problems. This being from someone whose had Dragon Age crash 3 times in the past couple hours. Fallout 3, how game makers get away with making such incredibly buggy software and being successful and not fixing it I really don't know.
 
Last edited:
Going to check this when I get home. I have severe lag spikes when I play mass effect 2. Never been able to solve it. Only when I move into new enviroments or spin the camera. Made the game unplayable really
 
isn't it better to have cores parked?

at least for LGA1156 i7 8x0 CPU and future Sandybridge. where they have/will have more advanced turbo mode. but the turbo mode only works when most other cores are idle.

eg, i have my CPU currently setup at 3.5GHz, and allow it to turbo boost up to 4.2GHz. for single threaded games, surely that turbo boost will be advantagous when other cores are parked (aka, other cores at C-state C6 thus allowing the single core to be over volted and turbo boosted)
 
drunkenmaster said:
Different kind of stuttering at a guess. Things like, I dunno for me Just cause 2, no idea why but with vsync enabled, stand perfectly still I'm rock solid 60fps, generally move in one direction, solid fps, turn a tiny amount of turn around completely and it jerks down to low 20's and slowly (over a couple seconds) gets back to 60fps.
Click to expand...

I get this, but with vsync disabled instead. Although my setup isn't as epic as yours to be able to manage 100-120fps without vsync on, I still manage at least 60fps when the game's frame rate doesn't decide to plummet for seemingly no apparent reason.

Like you mention, I can be standing still and facing one direction and be getting rock-solid fps, but if I turn just the slightest my fps goes down to about 45. Also, whenever this happens, my fps never goes below 45, so it's almost like my fps is being capped when it does happen. I've also experienced other instances where I'm getting about 45-55fps (like in big firefights), but then I change direction slightly and it goes down to 30fps, yet never below (as if it's being capped again).

I was talking to some guys in my class at college about this, and one of them thought something could be being rendered that's taking a chunk out of my frame rate, and I guess that could mean it's an issue with the game engine.

Alternatively, I suppose some form of vsync could somehow be enabled despite it saying it's disabled in the graphics menu (the only reason I say this is because of the annoying way Just Cause 2's vsync works).
 
Last edited:
DUFF-MAN > Would it be possible for you to run some tests with with your latency checker program please? (hopefully for single (and maybe crossfire/SLI setups):-

HPET ON
HPET ON + CCC process closed down
HPET ON + CCC process closed down + disable MMCS throttling feature

HPET OFF
HPET OFF + CCC process closed down
HPET OFF + CCC process closed down + disable MMCS throttling feature

MMCS: http://forums.steampowered.com/forums/showthread.php?t=1043281

Obviously, if anyone can help run some tests then great :-) I don't believe I have the HPET option in my BIOS to help test :-(

Just to relist, it was a great link that kitch9 showed earlier: http://forums.guru3d.com/showthread.php?t=332395 (same link as kitch just the full thread from post #1 instead of 'just' post number #2 :-)).
 
Last edited:
Wondered if it would stop the odd brief pause in some games so was gonna try the ops link but can't seem to find the key mentioned in regedit (via edit/search). Win7 64 with phenom quad 965. Is it just as relevant to the phenoms? Ta for the link op.

It's ok found it manually. (search just comes up with finished searching but didn't display a result)
 
Last edited:
Scougar said:
DUFF-MAN > Would it be possible for you to run some tests with with your latency checker program please? (hopefully for single (and maybe crossfire/SLI setups):-
Click to expand...

Sure... Will have to be later tonight (when I get home), but I can run a couple of examples with and without HPET / MMCSS. I have a GTX480 SLI setup, so I can run single-GPU and SLI configs. Obviously can't test crossfire...

I should point out though that the tool I have is designed for testing framerate irregularity from frame-to-frame ('microstutter'). Skipping and other "macro" stuttering effects that only occur every few seconds won't really affect the result... The algorithm ignores the largest 2% of frametime variations specifically to avoid these kinds of 'occasional' phenomena skewing the microstutter results.
 
Thanks, very much appreciate you being able to try this out.

BTW... I did see one reference possibly suggesting that Nvidia doesn't use HPET (they may use the system timestamp instead), it's ATI that uses (HPET), 'possibly' why they seem to (which i'm sure you've referenced in the past) have higher irregularity?

Regardless, it would be interesting to see if you do get more 'regular' frame results.

Perhaps an ATI crossfire user would be willing to help if they have the HPET option in their BIOS?
 
Scougar said:
BTW... I did see one reference possibly suggesting that Nvidia doesn't use HPET (they may use the system timestamp instead), it's ATI that uses (HPET), 'possibly' why they seem to (which i'm sure you've referenced in the past) have higher irregularity?
Click to expand...

It's possible :) If someone with a crossfire setup wants to test it out, then I'd certainly be interested to find out if it makes a difference.

I did find that ATI setups tended to have a slightly higher irregularity than the nvidia setups, but the difference is pretty marginal. Without running a series of tests in a controlled environment I'm not sure I can say there is a significant difference between the two setups. The amount of microstutter tends to vary strongly from application to application, and from setup to setup, so without buying a load of hardware myself it's hard to be precise about whether crossfire is worse/better than SLI overall.
 
Thanks for the suggestion. :)

Mine was set to 32 bit in the bios, anyway disabled it and over a period of time I will get to see if there are any improvements.
 
If you look at the link kitch9 put up (and the full link of it from me), you will see there is a measurement tool for latency. That might be useful for you arc :-)

Also.. try closing CCC.EXE and MOM.EXE before running games to check for improvements :-)
 
Duff-Man> Realised I had forgotten that Raven did actually say it sorted his problem out.. so there may be some benefit for Nvidia as well.

OP> Sorry if this has gone off on a tangent! Original post was about core parking and how to disable it!!!
 
Turning off HPET completely screws up Black Ops for me - granted it a game issue rather than a system one - but be warned anyone using APIC timer over HPET your likely to get guns "jamming" constantly due to timing issues in the game.

Also I found that 2 out of 3 PCs run into stability issues with CPU parking disabled (even non-overclocked and systems that pass stress tests with CPU parking disabled), so if you get applications unexpectedly crashing after changing that setting thats likely to be the cause.

I haven't tested it extensively for performance, but Black Ops - which has performance issues - did get a ~30-40% performance gain disabling CPU parking and was much smoother, other games tho I didn't notice much if any difference.
 
Last edited:
Scougar said:
Duff-Man> Realised I had forgotten that Raven did actually say it sorted his problem out.. so there may be some benefit for Nvidia as well.

OP> Sorry if this has gone off on a tangent! Original post was about core parking and how to disable it!!!
Click to expand...
It's all good, anything where people will benefit from the info works in my book :D
 
You must log in or register to reply here.
Back
Top Bottom