Ordering a new card tomorrow...

[Maverick123w]

Yeah, but the 4x0 series seems to have much higher mins in many games out there which is pretty important to me.

Couldn't order a card yesterday :/ both my store suppliers won't have what I wanted back in stock till tomorrow.

[Blacklash]

Ya' I have to OC my cards to at least 960 on the core to match lightly OCed 470 SLi mins. Even then sometimes I'm 4FPS behind in mins. Below is an OCed 470 beating a 5970 by 5FPS in mins @ 1920x1200 with flippin' 8xAA.

http://www.tweaktown.com/articles/32...d/index13.html

The 5870 in the review can't hit 20 when the stock and OCed 470 are doing 34 and 39 respectively. Definitely poke through the reviews out there and pay attention to mins with AA on. For example, I'd definitely rather 34/55 over 20/70 myself.

[Toss3]

Originally Posted by Blacklash

Ya' I have to OC my cards to at least 960 on the core to match lightly OCed 470 SLi mins. Even then sometimes I'm 4FPS behind in mins. Below is an OCed 470 beating a 5970 by 5FPS in mins @ 1920x1200 with flippin' 8xAA.

http://www.tweaktown.com/articles/32...d/index13.html

The 5870 in the review can't hit 20 when the stock and OCed 470 are doing 34 and 39 respectively. Definitely poke through the reviews out there and pay attention to mins with AA on. For example, I'd definitely rather 34/55 over 20/70 myself.

Are you honestly basing your opinion on Far Cry 2? Besides min fps without a graph to accompany it means nothing. A small stutter lasting far less than one second could bring the min fps value down without affecting gameplay at all.
A fair review would be comparing an overclocked 5850 to a similarly overclocked GTX 470 with a % value for <20 fps.

[Blacklash]

Read Hardware Canucks conclusion. It says what quite a few people are saying.

http://www.hardwarecanucks.com/forum...review-18.html

"Luckily for NVIDIA, we didn’t stop at the average framerates and included (as usual) minimums as well. In this category the GTX 470 SLI is still head and shoulders above the competition when IQ settings are pushed since it doesn’t exhibit the numerous dips in performance that characterize the Crossfire experience. I don’t know about you but I would much rather have a slightly lower average framerates if it means my system won’t bog down when I’m fighting for my life in a game. Basically, a fluid gaming experience is what SLI is ideal for…or at least until ATI can figure out a way to bring up their minimum framerates in some games."

For people that care about strong mins overclocked 470 SLi is going to turn out to be a winner. Yes that reviewer did praise 5850 Crossfire and he also mentioned it having more harsh FPS dips and lower mins. Again, IMO 34/55 is more desirable than 20/70.

[Toss3]

Originally Posted by Blacklash

Read Hardware Canucks conclusion. It says what quite a few people are saying.

http://www.hardwarecanucks.com/forum...review-18.html

"Luckily for NVIDIA, we didn’t stop at the average framerates and included (as usual) minimums as well. In this category the GTX 470 SLI is still head and shoulders above the competition when IQ settings are pushed since it doesn’t exhibit the numerous dips in performance that characterize the Crossfire experience. I don’t know about you but I would much rather have a slightly lower average framerates if it means my system won’t bog down when I’m fighting for my life in a game. Basically, a fluid gaming experience is what SLI is ideal for…or at least until ATI can figure out a way to bring up their minimum framerates in some games."

For people that care about strong mins overclocked 470 SLi is going to turn out to be a winner. Yes that reviewer did praise 5850 Crossfire and he also mentioned it having more harsh FPS dips and lower mins. Again, IMO 34/55 is more desirable than 20/70.

Weren't we discussing single cards here? If you plan on going multi-gpu and don't mind spending 150€ more than yeah sure GTX 470 sli is going to be better than two 5850s, but only barely when OC'ed to the max. Plus GTX 470 sli could require a new PSU if you don't have an 800w< one from the start(and even that would be pushing it).

Would really like to see a max overclocked 5850 vs max overclocked GTX 470(bet they'd be pretty close).

Can't understand why [H] didn't do a proper review.

[Blacklash]

As far as FC2 goes in the link I provided, I guess my question would be why is the 5870 only 6FPS behind the stock 470 in averages but 15 in mins?

I only mentioned the 5970 because I noticed the stock 470 equals its mins and passes them when OCed. The stock 5970 beats the stock 470 by 25FPS in averages and by 15 when the 470 is OCed. I was also surprised that was with 8xAA which is supposed to the 5800s strong point.

Yes, I know it's FC2 and it's not like the 5870 and 5970 are putting up poor averages. They are not.

[Slytat]

Originally Posted by Blacklash

As far as FC2 goes in the link I provided, I guess my question would be why is the 5870 only 6FPS behind the stock 470 in averages but 15 in mins?

It's something I noticed immediately with the 480s, they just feel more responsive overall than the 5870s do and Im sure this has to do with a better min FPS.

Again, that's not to say ATI is bad or anything, just that I definitely prefer the 480s to the 5870s.

[brunner]

Originally Posted by Blacklash

For people that care about strong mins overclocked 470 SLi is going to turn out to be a winner. Yes that reviewer did praise 5850 Crossfire and he also mentioned it having more harsh FPS dips and lower mins. Again, IMO 34/55 is more desirable than 20/70.

To be fair, the gap of 7 to 14 percent at 1920 isn't as big as it seems because the 5850 CF lost big in 1 or 2 games where it still did fairly well. The results are much closer if you take away AVP which all ATI cards seem to have problems with.

[Toss3]

Originally Posted by Blacklash

As far as FC2 goes in the link I provided, I guess my question would be why is the 5870 only 6FPS behind the stock 470 in averages but 15 in mins?

I only mentioned the 5970 because I noticed the stock 470 equals its mins and passes them when OCed. The stock 5970 beats the stock 470 by 25FPS in averages and by 15 when the 470 is OCed. I was also surprised that was with 8xAA which is supposed to the 5800s strong point.

Far Cry 2 is an nvidia game and as such will run better on Nvidia cards. They all run it good enough though so including it in a benchmark is pretty much useless.
Games that most people play and care about: Metro2033, BF BC2, AvP, Crysis Warhead(just for fun), Stalker COP, Dirt 2, ME 2, ARMA 2 and Just Cause 2.

How far can the GTX 470 be overclocked at stock voltages?

Speaking of dual cards - that imho is another plus for the 5850 as the OP only has a 750w PSU so he wouldn't be able to add a second card later without upgrading his PSU.

[Blacklash]

STALKER CoP is a "good" title for ATi. Compare my results to this fellow with GTX 470 Tri-SLi. My cards were @ 960|1265. I'm not even sure he was overclocking. If I drop my core to 900 my mins fall to 32 for the first few tests and 26 for the last.

http://www.nvnews.net/vbulletin/show...2&postcount=45

Settings


Results


I blast him in averages in the first three tests, now look at my mins vs his. Notice a trend?

5850 Crossfire is a darn good value and FPS min junkies should go with 470 SLi.

[Toss3]

How well does triple crossfire scale compared to triple SLI? This is really derailing the thread though as you're comparing three cards all of a sudden(and not even fairly as his 980x is clocked at 4,3Ghz).

[Blacklash]

Originally Posted by Toss3

How well does triple crossfire scale compared to triple SLI? This is really derailing the thread though as you're comparing three cards all of a sudden(and not even fairly as his 980x is clocked at 4,3Ghz).

Very well, I will be quiet after this.

My point in all of this is nVidia seem to have the upper hand well with mins when talking about single card, or any version of multi-GPU with AA enabled.

BTW my CPU didn't stop me from putting up a 96FPS avg in test one

Folks may discount all the results that are cropping up if they like. My STALKER results are not an anomaly when matched with others out there.

BTW 2 , same test no AA. If my CPU were the culprit I doubt my mins would be jumping up by 20FPS or greater without AA.