ATi's CCC: Not the greatest but not that bad either.

[xtknight]

Originally posted by: beggerking
Yes. I'm talking about the performance degradation of paging in / out CCC, then paging in game.

Yes, that will be minor when it does occur, and will only occur if the physical RAM is full. It's not something to worry about IMO. And the paging in of the game will occur regardless of ATI CCC (obviously).

If CCC doesn't do anything in the background, then we don't need it stay in resident. Save on paging.

That's right. Kill all CLI.exe processes while gaming if you want to avoid the ever-so-minor hit.

Performance degradation depends on how much physical memory you have, the size of the game, and if there is anything else actively running in the background.

In terms of memory, that's correct. But like I've said, the background things will be paged out in a short about of time to avoid degredations down the road so they effectively don't hurt anything unless the page file is full. And when the physical RAM is full, they have to be paged out which does not take very long. Just clarifying.

 

[kmmatney]

I'm happy with the bang for buck that my ATI card gives me, but NVidia simply makes better drivers. Even when I don't have CCC running, I still have 2 CLI.EXE entries in task manager.

Also, my card cannot scale 1:1 properly. It can only scale with black bars if one side uses the entire screen length or width. It cannot do arbitrary 1:1 scaling, say like 1024 x 768. My old NVidia 6600GT could do that just fine.

I'm now uninstalling CCC. I have to say it's pretty bad for what it does - took me a long time to find the pixel scaling options, and then I found out they don't work. I'm giving ATI Tray Tools a go.

 

[xtknight]

Originally posted by: kmmatney
I'm now uninstalling CCC. I have to say it's pretty bad for what it does - took me a long time to find the pixel scaling options, and then I found out they don't work. I'm giving ATI Tray Tools a go.

Yeah, that subpar software isn't worth its minor hit. ATI Tray Tools is superior in EVERY way from what I've seen.

 

[kmmatney]

ATI Tray cools is very nice. There's a bit of a learning curve, as there are so many options, but it wasn't hard to find and adjust the important settings. I didn't realize it had advanced overclocking built-in and gpu temperature readings. It even has motherboard voltage and temperature readings! All with a < 1 MB install, and taking 3 MB of memory!

The game profile feature is very nice - I have max overclock applied to some games, and mild overclock to others.

The only thing I wish ATI Tray tools had is maybe one large screen with all settings shows at once, with tabs or something (making it a little like CCC). However its pretty easy to get used to finding things in the tray menu.

Glad to get rid of CCC, and use something that's better.

 

[5150Joker]

Originally posted by: kmmatney
ATI Tray cools is very nice. There's a bit of a learning curve, as there are so many options, but it wasn't hard to find and adjust the important settings. I didn't realize it had advanced overclocking built-in and gpu temperature readings. It even has motherboard voltage and temperature readings! All with a < 1 MB install, and taking 3 MB of memory!

The game profile feature is very nice - I have max overclock applied to some games, and mild overclock to others.

The only thing I wish ATI Tray tools had is maybe one large screen with all settings shows at once, with tabs or something (making it a little like CCC). However its pretty easy to get used to finding things in the tray menu.

Glad to get rid of CCC, and use something that's better.

That's the reason ATT is probably so low in memory consumption, because it doesn't have a single GUI that has all the tabs in place. Either way, yes ATT is superior to CCC and I never disagreed with that assertion. Just that CCC isn't as horrible as some made it out.

 

[beggerking]

okay. minor hits on performance sums up to major hits, thats what I'm talking about. well , just IMO its a bad program that eats performance. just my 2 cent.

 

[Ackmed]

There is no proof of any performance hit, major or minor. There is no proof that it "eats performance". As I said, run some benchmarks with it, and without it. Until someone does, its all speculation.

 

[BroadbandGamer]

Originally posted by: kmmatney
I'm happy with the bang for buck that my ATI card gives me, but NVidia simply makes better drivers. Even when I don't have CCC running, I still have 2 CLI.EXE entries in task manager.

Also, my card cannot scale 1:1 properly. It can only scale with black bars if one side uses the entire screen length or width. It cannot do arbitrary 1:1 scaling, say like 1024 x 768. My old NVidia 6600GT could do that just fine.

I'm now uninstalling CCC. I have to say it's pretty bad for what it does - took me a long time to find the pixel scaling options, and then I found out they don't work. I'm giving ATI Tray Tools a go.

Let me know if ATT lets you turn off the scaling with ALL resolutions.

 

[chinkgai]

Originally posted by: Ackmed
There is no proof of any performance hit, major or minor. There is no proof that it "eats performance". As I said, run some benchmarks with it, and without it. Until someone does, its all speculation.

i lose about 70mb of RAM and about 10-20 pts in 3dmark06 with ccc than with drivers only

 

[beggerking]

Originally posted by: chinkgai

Originally posted by: Ackmed
There is no proof of any performance hit, major or minor. There is no proof that it "eats performance". As I said, run some benchmarks with it, and without it. Until someone does, its all speculation.

i lose about 70mb of RAM and about 10-20 pts in 3dmark06 with ccc than with drivers only

Thank you for backing me up and ending this long, awful argument.

 

[beggerking]

Originally posted by: xtknight


Performance degradation depends on how much physical memory you have, the size of the game, and if there is anything else actively running in the background.

In terms of memory, that's correct. But like I've said, the background things will be paged out in a short about of time to avoid degredations down the road so they effectively don't hurt anything unless the page file is full. And when the physical RAM is full, they have to be paged out which does not take very long. Just clarifying.[/quote]

doesn't take long for 1 operation. it adds up as countless operations are done every (human) sec.

 

[chinkgai]

i back no one in this argument...just stating my observations..

if you have plenty of ram and dont care for a tiny performance hit and would like the funtionality provided by ccc, by all means use it

all else, use drivers only or drivers only and atitraytools or rivatuner

i personally dont use any functionality provided by the ccc so i run drivers only...



one other thing i noticed is windows loads slightly slower with ccc

my system comprises of an x1800xt and 2gb of ram and dual core for those who are wondering.

 

[Zstream]

10-20 point difference shows nothing, if you run the same test 3 times you will get ranges from 50-60 point difference everytime you run 3dmark. So I think you are full of it tbh or very unfamilar with 3dmark

 

[chinkgai]

Originally posted by: Zstream
10-20 point difference shows nothing, if you run the same test 3 times you will get ranges from 50-60 point difference everytime you run 3dmark. So I think you are full of it tbh or very unfamilar with 3dmark

lollerskates at ur ASSumption

im a 3dmark whore and have every futuremark benchmark installed on my computer.


if you know anything about the newer 3dmarks (meaning 05, 06) they are much more "accurate" than the previous versions since they are more videocard limited than cpu limited. 3dmark01 varies much more wildly than 05 or 06


and yes i ran them it more than once u noob, i ran it three times each and 06 is never 50-60 points off, in my system anyways.

so maybe u should try rather than raising the bs flag with ur ASSumptions. thx.

 

[Munky]

Does ATT work on the x1900's? I'd prefer to keep using it instead of switching to CCC.

 

[ArchAngel777]

ATI X1900XTX = 5147 3DMarks

CCC W/ ATI X1900XTX = 5127 3DMarks

Performance loss = 0.39%

Well within the margin of error with ANY benchmark... Not to mention insignficant.

I cannot imagine anyone wanting to run CCC with that sort of performance degredation!

 

[Ackmed]

^ yep, 10-20 points is negligible. Thinking it causes a noticable performance loss is pretty silly. And caring about 3dmark says is too, but thats another topic.

 

[chinkgai]

well, u forgot to mention that windows loads more slowly and it takes 70mb more ram, both of which is unneccesary if you dont use any ccc functions anyways

i, for one, do not use any ccc functions

 

[ArchAngel777]

Originally posted by: Ackmed
^ yep, 10-20 points is negligible. Thinking it causes a noticable performance loss is pretty silly. And caring about 3dmark says is too, but thats another topic.

HAHA, totally right... That program isn't even good for a tech demo anymore.

 

[chinkgai]

there will ALWAYS be the few that scoff at something (3dmark) even though they run it themselves. but hey, its just for kicks right?

 

[ArchAngel777]

Originally posted by: chinkgai
there will ALWAYS be the few that scoff at something (3dmark) even though they run it themselves. but hey, its just for kicks right?

Few? Last I checked no one took that benchmark seriously. I would say it is the minority who believe that 3DMark = Real World Performance.

 

[NoStateofMind]

Originally posted by: ArchAngel777

Originally posted by: chinkgai
there will ALWAYS be the few that scoff at something (3dmark) even though they run it themselves. but hey, its just for kicks right?

Few? Last I checked no one took that benchmark seriously. I would say it is the minority who believe that 3DMark = Real World Performance.

Exactly! 3Dmark is just a way of saying "e-penile implant successful", REAL WORLD (without the implant) is so much more important.

*Surgeon removes implant*

 

[ArchAngel777]

Originally posted by: PC Surgeon

Originally posted by: ArchAngel777

Originally posted by: chinkgai
there will ALWAYS be the few that scoff at something (3dmark) even though they run it themselves. but hey, its just for kicks right?

Few? Last I checked no one took that benchmark seriously. I would say it is the minority who believe that 3DMark = Real World Performance.

Exactly! 3Dmark is just a way of saying "e-penile implant successful", REAL WORLD (without the implant) is so much more important.

*Surgeon removes implant*

LOL, good one...

 

[chinkgai]

rolfcopter. you guys are tooo funny.


i run it for fun and comparitive purposes. i do not associate it in any way with "real world" performance as that is highly game dependent and video card (mfg) dependent in some situations. (for example, a regular gtx spanks my x1800xt in aoe3)


i believe many people, not all, but many run 3dmark because it is a standard generic benchmark that everyone runs at the same settings for an easy comparison. how many games can you say that for? probably not many.

to compare "real world" performance, you would have to run a benchmark in the same game, at the same resolution, and even at the same part of the game. how do u do this? run fraps right? but how do u run every test between users and have equal testing grounds? and how many people will have the same games?

the solution is a generic benchmark such as this, isnt it now?

you guys humor me to no end, always barking that "real world" performance and epenis crap as if everyone does it for braggin rights and no one does it for comparitive purposes.

 

[ArchAngel777]

Originally posted by: chinkgai
rolfcopter. you guys are tooo funny.


i run it for fun and comparitive purposes. i do not associate it in any way with "real world" performance as that is highly game dependent and video card (mfg) dependent in some situations. (for example, a regular gtx spanks my x1800xt in aoe3)


i believe many people, not all, but many run 3dmark because it is a standard generic benchmark that everyone runs at the same settings for an easy comparison. how many games can you say that for? probably not many.

to compare "real world" performance, you would have to run a benchmark in the same game, at the same resolution, and even at the same part of the game. how do u do this? run fraps right? but how do u run every test between users and have equal testing grounds? and how many people will have the same games?

the solution is a generic benchmark such as this, isnt it now?

you guys humor me to no end, always barking that "real world" performance and epenis crap as if everyone does it for braggin rights and no one does it for comparitive purposes.

What is comparative about it though? I suppose you could argue that you can compare the same card with different clocks to see roughly how much performance increased. But even then, you can use a real world benchmark that is much better off and showing tangible performance increases.

I think some of the orgional 3DMarks were fun and neat but I never took their performance numbers seriously. Now, they fail to impress even a little, in my opinion.

Besides, it isn't about humoring you and the post wasn't directed at you anyway. Anyone who deems 3DMark to be a worthy benchmark that is on par with something like Quake4, HL2 or F.E.A.R. is who we are referring too.

Many cards have done very well in 3DMark only to find out they can't play the newest game, even though 3DMark told them it was the future! But who is to blame, 3DMark or those that believe it?