Sign in with
Sign up | Sign in
Your question

New Cat 8.10s looking good!!

Last response: in Graphics & Displays
Share
a b U Graphics card
October 1, 2008 4:56:39 PM

Heres some info with the 8.10 betas http://translate.google.com/translate?hl=en&u=http%3A%2... Looks like its averaging 10-20% better from the old drivers (8.7) in minimum fps, and, the best part is a manual fan control in the CCC

More about : cat 10s good

a b U Graphics card
October 1, 2008 5:17:32 PM

Youre sneaky heheh. This is a good site. This sites findings on other things have held their own on other forums, not just respected at TPU, but ES as well. So, I think youll find their findings consistant
Related resources
a b U Graphics card
a b \ Driver
October 1, 2008 5:27:12 PM

Nice changes - Eagerly awaiting when I can download the WHQL version and drop 8.9.
October 1, 2008 8:26:58 PM

Hmmm, this could be the one that everyone has been waiting for. Would be nice if they unlocked the GPU clock limit so we could get over 790 easier.
a b U Graphics card
October 1, 2008 8:44:33 PM

it be nice if I could have more stability with my 4850 @700core :p  I have to play on 680 to have it 100% stable......
October 1, 2008 8:48:48 PM

alvine said:
it be nice if I could have more stability with my 4850 @700core :p  I have to play on 680 to have it 100% stable......


Use this to test for stability: :p 
http://www.ozone3d.net/benchmarks/fur/
October 1, 2008 9:23:33 PM

Niicceeee. And ditto on what spinach said, auto-tune puts it at 790 every time, putting it at 800 on my own in the CCC or rivatuner causes instability.
October 1, 2008 9:31:42 PM

sweet. So far I have seen no difference from going from 8.6 to 8.9, I hope this one will really increase FPS this time. Even if it is up by 10%
a b U Graphics card
October 1, 2008 11:21:58 PM

A little off-topic:

Shouldn't be Cat 9.0 or something of the like? XD

I mean, 8.10 = 8.1, lol.

Esop! XD
a b U Graphics card
October 1, 2008 11:37:56 PM

It generally runs by months. So a 8.12 will most likely be seen, as they are released monthly. I think 1 time they actually diverted from this, and added an extra driver. But yeah, 1 a month
a b U Graphics card
October 2, 2008 3:57:20 AM

JAYDEEJOHN said:
It generally runs by months. So a 8.12 will most likely be seen, as they are released monthly. I think 1 time they actually diverted from this, and added an extra driver. But yeah, 1 a month


=o

Didn't know that, lol.

Esop! XD
October 2, 2008 4:48:29 AM

Will get it just for fan control.
a c 130 U Graphics card
October 2, 2008 7:13:43 AM


As San Pedro says worth getting for the fan control alone. Didnt read the whole article but will later, I got the impression from the charts that either the support has dropped off for the 3 series cards, just seems they are concentrating on getting performance out of teh 4 series or they are as good as they are going to get and we should just get these drivers for the fan support and call it a night. Any views on this thought ?
@ JDJ i like the site its easy to see whats going on and i like the multi testing on differant cards and OS :) 

Mactronix
a b U Graphics card
October 2, 2008 1:45:17 PM

Yea, this site is innovative, more than any Ive come accross. All sites (ggod ones) do things like they do, but not as often. I wish theyd just do larger looks, whether its cards,games or cpus, in their articles. I guess I like it so much, I just want more heheh
a b U Graphics card
October 2, 2008 2:08:27 PM

I wonder, how come CoD4 get's such a jump in FPS and the others don't... What's the "trick" under his engine?! >_<

Wich they could share some with Crytech, lol.

And yeah, being in german it's easy to understand. Good thing we use the same "numbers", lol. Thanks JDJ :p 

Esop!
a b U Graphics card
October 2, 2008 2:43:19 PM

YW. Itll come out in English later, more than likely, they have translators, but its first in German
October 2, 2008 3:49:13 PM

Yuka said:
I wonder, how come CoD4 get's such a jump in FPS and the others don't... What's the "trick" under his engine?! >_<

Wich they could share some with Crytech, lol.



CryEngine2 is just too much for current GPU architecture to handle. No driver update can help it. I am hoping that DX11 being multithreading will really up the anti so games as demanding as Crysis will run like COD4. It would be interesting to see a CryEngine for DX11. It would be nice however to see ATI man up to the crappy xfire support for Crysis. SLI is killing them at the moment.
a b U Graphics card
October 2, 2008 4:54:50 PM

I don't know... I actually have my doubts on the CryEngine2 and it's "optimizations". Seems like they still load a lot of things to the CPU IMO, but it could be, like you say, that cards still don't use/have the technology Crytech wants to use in it's engine. It could also be due to it's "open" theme without much thinking and just throwing all out at once.

Well, i read the whole credits of Crysis and there where 5 people from nVidia and 3 from ATI if i recall correctly, so i think this game shouldn't be so biased towards nVidia, but it certainly feels like it. So improving CF on it won't do much IMO, since 1 card config might not even get everything it should from the game. Ahh, we gotta love the fud, lol.

Esop!
October 2, 2008 9:10:26 PM

Its just a tech "Demo", nothing more. Thats all Crytek games are, Gameplay is the last thing on their mind.
a b U Graphics card
October 2, 2008 9:58:38 PM

It's funny peple have issues with the site, it and ComputerBase are like the 2 biggest German language gaming hardware sites.

Along with Chip.de and 3dcenter.org I usually check out PCGH and ComputerBase.
October 2, 2008 10:09:43 PM

spathotan said:
Its just a tech "Demo", nothing more. Thats all Crytek games are, Gameplay is the last thing on their mind.


This is off topic but I'm compelled to disagree. Warhead was fun. I'm playing through a second time without using stealth to add more challenge. I've rarely played a game twice.

On topic the only thing I can see better about the driver is the fan control for those who keep asking how to adjust it. I'll wait for the release before I try it but what I really want to see is the 2d mode at startup problem fixed.


October 2, 2008 10:15:18 PM

spathotan said:
Its just a tech "Demo", nothing more. Thats all Crytek games are, Gameplay is the last thing on their mind.


No1sFanboy said:
This is off topic but I'm compelled to disagree. Warhead was fun. I'm playing through a second time without using stealth to add more challenge. I've rarely played a game twice.

On topic the only thing I can see better about the driver is the fan control for those who keep asking how to adjust it. I'll wait for the release before I try it but what I really want to see is the 2d mode at startup problem fixed.


Meh, it's just a matter of opinion. I liked Crysis, Warhead, not so much. It's just too much shooting and too little strategy. :p 

And the fan works fine as is. 4850/70 may run hot, but they also tolerates much higher temperatures compared to other gpus. They're known to hit 100C without losing stability or even artifacting. The low fan speed is supposed to keep noise low. Fears concerning high temperatures are unfounded.
a b U Graphics card
October 2, 2008 10:23:06 PM

I ran accross the fix earlier, you still need it Ill try to find it again.
October 2, 2008 10:43:24 PM

dagger said:

And the fan works fine as is. 4850/70 may run hot, but they also tolerates much higher temperatures compared to other gpus. They're known to hit 100C without losing stability or even artifacting. The low fan speed is supposed to keep noise low. Fears concerning high temperatures are unfounded.


These things at default settings idle at 80c and reach 100c under load. When you look further than the temp reported by CCC using Rivatuner or GPUz you see other parts of the card like shader core reporting temps another 7-11c higher.
Now the quote posted in the other thread where you kept trying to preach to leave the fan alone had ATI's Eric Demers saying the cards are rated to 105c. I think the cards are hitting these temps under normal operation. Time and time again the people who own these cards reporting that at 35-40% fan speed the card runs much cooler with reasonable acoustics. I don't understand why you insist on owners of these things simply letting them run hotter. I agree some people do overkill by trying to reach ambient with their gear but bordering on max is poor practice.
October 2, 2008 10:51:16 PM

35-40% is definetly the sweet spot as No1s says. Ive mentioned this more than a handful of times that its what I run mine at. 45-50% is where it starts to get out of hand as far as sound goes. I put it at 100% for **** and giggles and could hear it in the next room through the wall, doors shut. Im currently gathering funds for the Thermalright T-Rad and two 92mm silent fans to put it on (about $75 all together :(  ), this stock cooler has to go, im sick of this noise and sick of living on the edge of the thermal limit.
a b U Graphics card
October 2, 2008 11:21:11 PM

[brag] Damn, i love my dual slot fan Sapphire! [/brag]

Sorry, had to do it XD

And dagger, like i said somewhere... It's not just about the cards own temperature. If you have a component that can radiate 150ºc inside your case look around for a melt down, lol. Well, maybe not a melt down, but you know what i mean XD

These cards (BBA design for 4850's) really hurt the case temps inside, making everything hotter. So you might get a "silent card", but later on you'll have to add some fans and make holes (get me mah HAMMA'! lol) or even buying another cooler for the card.

Anyway, i measured no increase with mine, so i really thank Sapphire for their design, and actually the temp dropped compared to the X1800XL, lol.

I wonder if non-reference boards will still benefit from that in 8.10's CCC... I suppose they will, but on different RPMs. Did they do a % bar or RPM bar? lol

Esop!
a b U Graphics card
October 3, 2008 6:49:16 PM

The back plate radiation on these is nothing comapred to the them exhausting back into the rig.

The HD4870 could produce 120C backplate temp, and exhaust most heat out the case and still be less of an impact than an HD4850 having a 20C lower temp, yet most of the heat stays in the case.

That's part of the reason why it's not just about the temperature but the effectiveness of the cooling, and the HD4870 having cooling that may improve case airflow makes it's temps a non-issue..
a c 169 U Graphics card
a b \ Driver
October 3, 2008 7:04:21 PM

WOW! Very nice thanks jaydee for the link :) 
I hope i will get better FPS in STALKER and WARHEAD and of coures fan speed control :) 
a b U Graphics card
October 3, 2008 9:19:36 PM

I think the AA issue is fixed with these in Stalker, not sure tho
October 3, 2008 9:27:29 PM

True grape ape, at 45% speed the air out the back is blistering hot.
a b U Graphics card
October 4, 2008 5:52:20 PM

spathotan said:
True grape ape, at 45% speed the air out the back is blistering hot.


Wonder if ATI should promote their 4870's as "buy one and save on air conditioner in winter!", lol.

Esop! XD
a b U Graphics card
October 4, 2008 6:04:33 PM

The 4870 S'More edition heheh
a b U Graphics card
October 4, 2008 6:46:37 PM

Quote:
not that ridiculous, been a few years since i have had heating on in my room, comp does just fine. Now that it is coming to winter i will to have to have my window open all the time but to get airflow someties it will be.


Well, electrical heaters are 1.2W to 2kW, so... a 500W consuming Rig should be able to heat up a room in a little more time, lol.

Summer is the issue though >_<

Esop! XD
a b U Graphics card
October 4, 2008 7:05:51 PM

Something you guys seem to miss, the GTX280 obviously puts out more heat out the back, since more power consumption = more heat (almost all power is converted to heat in GPUs & CPUs [minor amount of radiation in other forms]), the big difference being the efficiency of the HSF. Even the GTX 260 consumes more power while playing, so overall both the GTX 280 and 260 would put more heat into your room than the HD4870.

http://www.xbitlabs.com/articles/video/display/force3d-...

Remember you move significantly more heated air at near the same temperature(K), then you're outputing significantly more heat into the room while the component temp may be lower.
October 4, 2008 7:13:03 PM

since the 4870s and 50s have soo much headroom would the 280 GTX also have as much?

Performance wise, can i expect 10-20% or not?
a b U Graphics card
October 4, 2008 7:36:43 PM

Ive heard the G280s are less apt to oc as well as their little brothers, but even the G260s are crap shoots, one will oc great, while another just so so.Since the 280 is the top card, its pushing the limits already, thus less headroom than the 260.
@ Ape, are the hsf's moving the air that slowly then? Thatd make it seem hotter...
I did some thinking on this, and Im wondering if the nVidia cards push out more air, since only maybe half actually goes out the back of the case. More going thru the card it would seem. And looking back at the 8800GT, and all the heat issues they had until they corrected their cooling solutions, it makes sense, as they couldnt do it the way ATI does and get away with it
a b U Graphics card
October 4, 2008 8:32:31 PM

JAYDEEJOHN said:

@ Ape, are the hsf's moving the air that slowly then? Thatd make it seem hotter...


I don't know for sure, but it seems like it. As you can see in the link the HD4870's HSF sounds quieter too, so RPM may be lower (remember the issue for many was ~20% fan speed being default), but the overall design may be different too, is it still a 'partial in / mostly out' airflow design on the nV's HSF assembly?

Quote:
I did some thinking on this, and Im wondering if the nVidia cards push out more air, since only maybe half actually goes out the back of the case. More going thru the card it would seem. And looking back at the 8800GT, and all the heat issues they had until they corrected their cooling solutions, it makes sense, as they couldnt do it the way ATI does and get away with it


Yeah and that's what I don't know, could be still a 70/30 design with some pouring back into the case. Same volume of air leaving the case but overall more cooling, but if some is coming back into the case that would be bad for inefficient cases, but fine on an OK case, but a concern to those here in this thread. It would also give the apperance of lower temps, but actually have less effective eco-systyem cooling solution.

The thing that get missed is the temp/heat distinction of course. It's like saying, what would you prefer to heat an entire house, a single flame of an Oxy-Acetylene/Propane torch @ 2700-3200c , or a nice wood-stove @ ~600c ? :sweat: 
October 4, 2008 8:39:58 PM

It might also have to do with smaller die size, which translates to heat production concentrated on a smaller area. And since the onboard heat sensor is right there... :p 

Also, not sure about gtx200s, but 8800gts reference dual slot heatsink is physically heavier than ATI counterparts.
a b U Graphics card
October 4, 2008 8:48:24 PM

I miss my fireplace. But yea, power in thermals out. But in a fair comparison, the 8800GT couldnt use the same cooling technique. Id still much rather have it out of the case. Just like a torch, Id rather use it outside too heheh, and leave the woodstove inside heheh
a b U Graphics card
October 4, 2008 8:54:44 PM

The die size is relative to heat found in one area, not total heat, just so no one thinks otherwise. It actually doesnt put out more heat, actually a little less, but from a smaller source, yes. I still think its design. 8800GT couldnt do it, they got too hot. I wouldnt put tape over my inside hsf vents, you wanna try dagger? heheh
a b U Graphics card
October 4, 2008 10:38:53 PM

dagger said:
It might also have to do with smaller die size, which translates to heat production concentrated on a smaller area. And since the onboard heat sensor is right there...


Which just supports the same idea further with regards to the effectiveness of the HSF, and I have already mentioned previously as well many times before with regards to surface area of the HSF interface with ever smaller chips.
Regardless of whether it's in die, on package or on PCB, it still needs to be calibrated so that the raw sensor info corresponds correctly to actual temps, and that's something that often has to be tweaked for Rivatuner and GPU-Z just like many other changes from model to model, it's just easier to fix.

Weight of the HSF doesn't mean much on it's own unless you know it's equal composition, like if it's pounds of steel, heavy motor and heavy housing instead of copper heatpiping and light motor with plastic housing. But from what we've seen there's alot of copper looking stuff in the GTX, the biggest question to me is whether it's copper coated aluminum like on the ATi, which is not as good a all copper, but it's lighter than both all copper and sufficient aluminum.

JAYDEEJOHN said:
The die size is relative to heat found in one area, not total heat, just so no one thinks otherwise.


Yeah dagger's sorta responding late to the previous discussion and not quoting, so it look like he's not quite getting it, it's just he's replying to earlier in the thread without attribution, which is annoying. :pfff: 

Quote:
It actually doesnt put out more heat, actually a little less, but from a smaller source, yes.


Exactly, and it does contribute to it as well, but whether it's interface size, composition or other characteristics of the HSF, the end result is it's not as effective at removing the heat generated by the chip as that on the GTX, however as long as the chip is stable, it doesn't really matter much other than showing a higher number, which would be like saying the HD4870 is better than the GTX280 because the number 4870>280, which is why the GF8800 was better than the HD2900 because once again 8800>2900.
a b U Graphics card
October 4, 2008 10:52:02 PM

LOL too true. And since nVidias TDP must actually be closer to their claim than ATIs, maybe thats why the concern for the 8800GT, as they say they cant run as hot on nVidia.
!