9800GX2 vs HD3870X2 full review!

review_hunter

New member
Finally a proper review is out, looks like 9800GX2 beat the HD3870X2 in almost every benchies! It's a pity that these early forceware drivers are buggy. Review

in English:

http://lly316.blogspot.com/2008/03/geforce-9800-gx2-vs-radeon-hd-3870-x2.html

in Chinese:

http://www.pconline.com.cn/diy/graphics/reviews/0803/1241871.html

Here are some screenies:

0112418710803129800gx23kq5.jpg


0212418710803129800gx23ip0.jpg


eb080480b1.jpg


0612418710803129800gx2ckg2.jpg


3fef7e9c9f.jpg
 
Has AA problems but they can easily be fixed. Forceware drivers have been particularly buggy with AA recently.
 
Its not bad, as expected really, 2 GT's in sli are faster than 2 3870's in CF, 2 GTs cost more...

With the 'new' gtx coming in at $300 (see the 9800gtx testing thread) i think 790i with 9800gtx sli is my next rig. maybe.

AMD have 'Thunder' which is set to shake things up again!

Not that impressed with anything atm tbh
 
name='Mr. Smith' said:
Its not bad, as expected really, 2 GT's in sli are faster than 2 3870's in CF, 2 GTs cost more...

With the 'new' gtx coming in at $300 (see the 9800gtx testing thread) i think 790i with 9800gtx sli is my next rig. maybe.

AMD have 'Thunder' which is set to shake things up again!

Not that impressed with anything atm tbh

AMD Thunder does sound interesting. Has Thunder got anything to do with the rumored RV770 with 800 shaders?
 
Hey, do read the review again, seems like the one in English, updated the review.

http://lly316.blogspot.com/2008/03/geforce-9800-gx2-vs-radeon-hd-3870-x2.html

I spotted these changes:

CoJ --> Although the 9800 GX2 is 1 fps faster than the AMD, but it was unable to run the benchmark with 4xAA turn on. It happened at the moment the benchmark begins, it bounced right back into desktop. Therefore, the score is recorded as zero.

UT3--> Correction:

Just like the previous game benchmarks, 9800 GX2 took the lead easily without AA.

But performance decreased drastically with 4xAA/16xAF turned on, resulting with a single digit fps on the average. Thus the score is once again recorded as zero. (There were no errors encountered here)

WIC--> Again, when AA is turned on for the GeForce 9800 GX2, an error message appears. Thus the score is recorded as zero.

Do tell me if you spotted other updates. ;)
 
I think the majority of people here can kick this/these reviews directly into touch.

Either that or explain to me how benching the cards at 2560x1600 is of any relevance ?

Think there are a handful of people here with 24" that do 1920x1200, but I would hazard mainly 20-22" 1680x1050 resolutions, and even a spread of 19"-20" 1280x1024 resolutions.

More 1024x768 users I`d imagine.

What is the point of beating a card, on paper @ 2560x1600, if you similarly blow at lower resolutions ? It`s not an automatic fact that a gfxcard will beat another at a lower resolution cos it beat it at the higher end.

And they use "realistic ingame worldinconflict" benches.... @ 2560x1600 - realistic ?
 
You`d be suprized. But what shouldn`t suprize u is they are not sporting as many 2560 screens, and the reviewer should know better.
 
15" FTW! I think the norm is 22" and high cost is 24", I mean even try SLI XFX XXX 8800 Ultra's can't handle crysis fully at 1920x1200. That is largely because of crap drivers though.
 
U know, I think 22" is becoming the biggest consideration for enthusiasts, that or 20" depending on the desk-sizes etc, cos frankly there isn`t that much in the price. Both ~usually~ 1680x1050 (which is why I like 20" tbh).

24" is the nice end, and a good few guyz have them here - there`s the quality step up.

Definately on my list of purchases this year. (but with two inputs for me)
 
24inch goodness woot woot!

20inch is probably the best atm though, nice performance/quality level for current cards.
 
Back
Top