9600GT SLI beats 3870X2 in a couple games (inq, expreview)

from what i can see form all of the reviews i have seen is that the HD 3870 and 9600gt perform very similar to each other. The cards seem to be mostly tied when gaming without AA (it bounces around lots form game to game) they 9600gt may have a TINY lead without AA and when you turn AA on the lead seems to grow a little bit. but overall they are pretty close.

so lets say performance is the same. it comes down to:
power consumption - relatively the same
heat - i would say 9600gt is a bit cooler
features - the HD3870 does have DX10.1 but that is useless for while. but overall HD3870 wins
overclocking - form what i have seen so far the 9600gt oc's like a mad man (past 830 core i believe) and considering shadder clocks scale higher than the HD3870 i would say the 9600 wins by a small margin
drivers - personally i don't like ATI's drivers and i prefer ntune to CCC overdrive
Multi GPU - again seems to be very similar depending on the game and settings
and finally price - Where i come form (Canada) the 9600gt is cheaper than the HD3870 by a good amount. the 9600gt can be had for $180 were the HD3870 is $200 (after the price drop) and if you go for multi GPU price thats a $40 savings. and $360 for a pair of 9600gt's is far cheaper than $450 for a HD3870X2

this is for Canada not europe and not the US (although they tend to similar) and form what i can tell prices in europe tend to favor ATI cards so in that case the Hd3870 would be the better choice.

for myself i would choose the 9600gt but both cards are good and are very competitive to each other.

Quote:

---

Originally Posted by Sr7

So whatever is *wrong* is only affecting ATI?

Logic Fail! :down:

What's wrong with using an SLI motherboard? Does the chipset have an "Auto-slowdown-ATI" feature in the BIOS?

---

My board actually has a "Ati detected, speedupmode enabled". It sets the PCI-E frequency to 125 MHz when and Ati graphicscard is detected ,-) (RD600).

This just proves that ATI needs to get CrossfireX drivers ironed out and needs to lower the price on their cards.

What's interesting is that if you go by Expreview numbers then 9600GT SLI is the fastest setup on the planet, faster than 8800GT SLI and 8800Ultra as well. Is that truly possible?

If so to hell with the next gen, let's all buy 9600Gt's and SLi them! :D

Only problem is how do I use EVGA stepup for my 8800GTX to get a 9600GT?

The launch has been delayed so many times... Maybe Nvidia realised they have spent the last year developing a card that can't actually match the Ati equivalent and are desperately trying to get performance up before they have to release it.

http://i11.photobucket.com/albums/a1...attachment.jpg
This pic is provided by another user using Vista and Maximus Extreme motherboard. He uses a Q6600 at 3.0 (I assume that a QX9650 would give a higher SM2.0 and SM3.0 score). But using what we have we know that QX9650 provides a score of 4509 looking at Hardware Secrets review. Using 3dmark06 calculator we get a score of 16292. This is not including the possible increase in SM2.0 and SM3.0 with a QX9650 from a Q6600. 16292 is by no means an accurate score but it does reflect the some increase when going from a X38 vs 780i. This in all shows that the score using a X2 with a 780i is lower then expected.

Quote:

---

Originally Posted by Eastcoasthandle

http://i11.photobucket.com/albums/a1...attachment.jpg
This pic is provided by another user using Vista and Maximus Extreme motherboard. He uses a Q6600 at 3.0 (I assume that a QX9650 would give a higher SM2.0 and SM3.0 score). But using what we have we know that QX9650 provides a score of 4509 looking at Hardware Secrets review. Using 3dmark06 calculator we get a score of 16292. This is not including the possible increase in SM2.0 and SM3.0 with a QX9650 from a Q6600. 16292 is by no means an accurate score but it does reflect the some increase when going from a X38 vs 780i. This in all shows that the score using a X2 with a 780i is lower then expected.

---

I really hope you're kididng me... 100 points is proof? Of what? Even if it is 100 points higher, that doesn't prove that that phenomenon only affects X2.