Catalyst 10.9 is out

[wolfram23]

Hmm so pretty minimal change to 10.9. Well, I did a comparison a while back, I found 10.6 gave me some decent FPS increases in a few games. I've always just used the latest drivers tho, haven't had any issues except the odd bad install which a proper driver sweep fixed.

Those STALKER results are definitely odd tho. I guess maybe has to do with no AA, but I'm running CF 5850s at 875/1200 or so in this test (I think this was with 10.5):

I'd make doubly certain that everything is set to "Use Application Settings" in CCC

 

[OvrClkr]

Hmm so pretty minimal change to 10.9. Well, I did a comparison a while back, I found 10.6 gave me some decent FPS increases in a few games. I've always just used the latest drivers tho, haven't had any issues except the odd bad install which a proper driver sweep fixed.

Those STALKER results are definitely odd tho. I guess maybe has to do with no AA, but I'm running CF 5850s at 875/1200 or so in this test (I think this was with 10.5):
http://i746.photobucket.com/albums/xx103/Wolfram23/5850 single vs CF/STALKER.jpg

I'd make doubly certain that everything is set to "Use Application Settings" in CCC

Nice chart !

Do you have any Dirt2/BC2 charts ?

 

[wolfram23]

I ran tests in AvP, Dirt 2, Crysis, FurMark, STALKER, and RE5
http://www.tomshardware.com/forum/289809-33-single-5850-furmark-crysis-stalker-dirt

 

[OvrClkr]

I ran tests in AvP, Dirt 2, Crysis, FurMark, STALKER, and RE5
http://www.tomshardware.com/forum/289809-33-single-5850-furmark-crysis-stalker-dirt

sweeeet 😉

TY!

 

[Doom3klr]

Im using 10.5s there pretty good how do they compare to the 10.4s?

 

[Mousemonkey]

Im using 10.5s there pretty good how do they compare to the 10.4s?

Why don't you just install the 10.4's and find out?

 

[Doom3klr]

You got a point there thanks.

 

[darkknightdude]

You must be the luckiest guy on earth, especially when there were no improvements whatsoever for those games you listed above.

In my last 15 years of playing FPS games I have never seen a driver give the user an extra 25 frame boost, man you got lucky!! :lol:

You obviously haven't tried 10.3 😀

 

[notty22]

I believe 5850 at 1000/1200 speed gets you performance equal to 5870.

Tech power up has done a lot of 58XX reviews, most times 5850@900-925 equals 5870@850 results. Not to argue with you. I think the consensus is about 8% difference because of the shaders.

 

[aznshinobi]

Actually Notty is right. When I use 1000/1205 clocks, they beat the standard score of a stock 5870. Not to say 3DMark Vantage IS ALWAYS the most accurate way to tell. But a stock 5870 gets a score of 17xxx I can't remember something like that. But my 5850 @ 1000/1205 is 18125. So clearly its better.

 

[Mousemonkey]

Actually Notty is right. When I use 1000/1205 clocks, they beat the standard score of a stock 5870. Not to say 3DMark Vantage IS ALWAYS the most accurate way to tell. But a stock 5870 gets a score of 17xxx I can't remember something like that. But my 5850 @ 1000/1205 is 18125. So clearly its better.

In the winter months I may have a go at screwing that out of my 8800's, would that make them the better choice?

 

[aznshinobi]

Wait what do you mean better choice? If you OC I mean sure XD All the extra core clocks count (don't know bout Nv cards never oc'd shader clocks) I mean from 950/1205 to 1000/1205 I got a noticeable increase in performance on 3DMark Vantage, however I didn't notice HUGE differences in games like Mafia 2 or anything else.

 

[Rustyy117]

Ok here is my finding.Done with 5850 and q9550 at stock at 1024-768 res.4 gb ddr2 800mhz ram

1.Re5 4xaa maxed setting variable benchmark.
Cat 10.3-126 flps
Cat 10.9-126flps

2.Hawx Maxed setting 8xaa
Cat 10.3-109
Cat 10.9-108

3.Just cause 2 maxed setting 8xcsaa desert sunrise
Cat 10.3 -86
Cat 10.9-84

4.Farcry 2 RANCH MEDIUM ULTRA SETTING 4X AA A.I DISABLED
cat 10.3-Average101 MIN 60 FLPS
Cat 10.9-Average 99 MIN 55FLPS

5.Mafia 2 maxed setting Apex physx of
Cat 10.3-74
Cat 10.9-70

6.Dirt2 ultra setting 8xaa
Cat 10.3- Average 87 MIN 72
Cat 10.9- Average 85 MIN 70

I am switching back to my everfaithful cat 10.3 never lets me down.

Not much point in me running those benchmarks for you as you've already done them Shame about the results though

@Ovrclkr, Yes there is, I think its called easytune, I don't have it installed atm though.

@Wolfram23, I think DX11 isn't working for me in stalker CoP and metro 2033 as I had to play both games in DX9/10 mode to be able to use a res higher than 1280 x 720, Something to do with me using a HDTV as a monitor....
It is weird though because I double checked and DX11 is deffinately enabled along with tesselation and the benchmark runs fine, Just my results are ridiculous.

 

[Rustyy117]

@Omg Mate what do you mean by 'shame about the results though'Are my flps low?

No, there not low, Just lower than 10.3

 

[Rustyy117]

Nothing wrong with the card, its my TV that I'm using as a monitor, DX11 works perfectly with my friends monitor in all games. Strange thing is, DX11 works fine in heaven benchmark at 1920 x 1080, but stalker and metro2033 don't.

Anyways I've got my eyes on a monitor, just trying to sell my TV atm.

 

[Rustyy117]

Did you get any framerate rate boost in re5,dirt2 benchmark with 10.9?And what happened to your sec 5850?

Not tried Dirt 2 or RE5 yet, I got rid of my ASUS card, too much hassle, some programs wouldn't work properly, got the odd BSOD + it was a horrible OCer, I think I'll just stick with one card, waiting on the 69xx series or NV's next gen.

 

[Maziar]

I'm going to download it(yes i haven't yet because of mind blowing internet speed :lol: ) and will post the results compared to Catalyst 10.8.

 

[Rustyy117]

Oh if anyone is having under/overscan issue's then the 10.9 hotfix fixes that (even though there for 48xx X2 cards).

Some people on another Forum saying that Crossfire profiles are improving performance even for single card users. I d/l the profiles and tried BFBC2 again and gained 1 FPS minimum, but lost 3 FPS Average lol.....

 

[Rustyy117]

You sold that card.I donot they are going to refund your money just becoz it was not overclocking.

Yeah I said " This just won't do, its unacceptable. This card will not overclock to 850Mhz Core without a bump in Voltage, I demand my money back now!" :lol:

 

[Rustyy117]

So they refunded your money.

No dude it was a joke, There not gonna refund someone because it doesn't overclock...

I sold the card to my bro ( didn't tell him it can't overclock for **** :lol: )

 

[sirbuster223]

When it comes to ATI and new cards, it usually takes them awhile to start rolling in with uber awesome drivers for there epic GPU. Definatly not as fast as nVidia, but they will get there with there 5000 series cards.

Who knows, maybe theyll do a better job with there new 6000 series coming out soon?
Lol

 

[Rustyy117]

When it comes to ATI and new cards, it usually takes them awhile to start rolling in with uber awesome drivers for there epic GPU. Definatly not as fast as nVidia, but they will get there with there 5000 series cards.

Who knows, maybe theyll do a better job with there new 6000 series coming out soon?
Lol

"ATI" and "uber awesome drivers" do not belong in the same sentense. Although to be fair these 10.9's have been an improvement on the last 5 months worth of drivers.

 

[wolfram23]

dipankar, the one issue I have with those results is that you can't rule out a CPU bottleneck due to the low resolution. I know that we're trying to see how the drivers change FPS but if it's a CPU bottleneck, you can't rule out that a background operation was underway causing you to lose a couple FPS. If we can guarantee it was GPU limited, then it'll be accurate. I'm surprised your RE5 results were 126, I know it's a CPU bottleneck on that one tho because I got 100 with a single 5850 and 127 with CF 5850s, and later on I re-ran it at 4ghz and got 150FPS (+/- 5)

 

[sirbuster223]

"ATI" and "uber awesome drivers" do not belong in the same sentense. Although to be fair these 10.9's have been an improvement on the last 5 months worth of drivers.

Your definitely right about my error omgitzfatal. They don't belong in the same sentence, I was just over emphasising on the word 'good'. Tehe.

But yes, I do agree with new 10.9 drivers being a improvement.

 

[stillerfan15]

Catalyst™ Control Center Version 2010.0406.2133.36843
I'm running this version of CCC for a 5570 with absolutley no issues. No reason to change things at this time IMHO.
Dave