i was wondering the same
Printable View
i was wondering the same
I guess someone didn't like the fact that OCed 3870 in crossfire scored 20 000 marks in 06 with 790 and Phenom.
nope
some Admin/mod just delet the old one because a lot of reviews as been released and there´s no need of multiple thread´s with same subject and they open a new HD3850/3870 Reviews thread
they did same thing with 8800GT thread
(forum = clean)
you can post the HD3870 CF 20K score again in this thread , what´s the problem?? :)
:up:
Below I took some snap shots similar to those found in the article. Let me make it clear it's impossible to get a photo of the "exact" same spot do to the movement of the camera in the demo. However, as you can see these photos dismiss any LOD excuse some may believe is happening.
http://i11.photobucket.com/albums/a1...rysisIQ3-1.png
Distance is further away then the SS in the article
http://i11.photobucket.com/albums/a1...rysisIQ4-1.png
This SS is a bit closer then the other. I would consider this in the "ball park" of the SS in the article.
However as you can see, the detail even at a distance further then the SS in the article the detail is similar. :yepp:
Sidenote:
SS taken using ATI video card.
Yeah, I figued that would be the case
NVIDIA boys keep saying cheats.... err bugs/drivers and so on, the evidences are there. Where is the LOD? :D
:p: :p:
8800GT only loud if fan + 50%-55% , under that its oK
mine its not very Hot , so i use it @ 30% (no noise at all) >> 45º Idle / 65º Full (atitool scan artifacts)
Idle:
http://aycu05.webshots.com/image/352...7846089_rs.jpg
Scan artifacts:
http://aycu34.webshots.com/image/354...2142256_rs.jpg
:up:
It looks like it was taken from the same horizontal plane or close to it. It may not, as some suggest, have been taken from the same elevation. Not that it matters at this point
I went ahead and studied the results of the TweakTown review again.
I looked at the results at the resolutions I would use and most are within a few percent of each other. There are a few that are >30% but that is on applications that are being rendered at 100 frames/sec already.
It seems to me that if some one were in the market to buy one of these parts, there is no real compelling reason to get the GT over the 3870. Yeah I understand the concept of more is better but personally I like more. . . quality. What I like less of is heat and noise not textures.
In Crysis though the difference is big but I'll wait for the 7.11 and deceide then
I'm pretty sure the IQ differences are in the 169.xx drivers which I'm not a big fan of. Perhaps 169.02 is quite good though but performance is a bit low for my liking. Personally I'd rather see Crysis with 167.26 driver or 163.75. Claiming the graphics card has worse IQ when it might be only a driver issue is lame as if it was hardware related it would be permanent, if software/driver related, no big deal and is fixable. So until you can't prove it's in the drivers or not, I wouldn't make a big noise about this as if it's hardware or software related has like a day and night difference and importance.
With NVIDIA you're far definitely not best off using the latest drivers all the time. I've been comparing nvidia drivers for many years (since I was using a Riva TNT2 M64 32MB card) and I'm into tweaking and hang on tweaking communities so I tell that by experience. Currently 167.26 is far the best driver for my setup but it won't necessarily mean it's a good driver for you.
Now compare that 2nd ss to my 8800GTS 320MB @ 169.09 drivers;
http://i17.tinypic.com/821m00m.jpg
Maybe different colors, maybe different gamma, possibly different settings. But the same IQ!
:up:
Can someone PLEASE post the video of Macci running over 20k in 06 with Phenom, 790FX, and 3870s. I did not get to see it before the thread was deleted.
Thanks.
I'm still betting my hard $$$ from Dennis review ... 20 k 3dmark06 in CF for 3870 ... That's fine with me !
Since the old thread was removed, Dennis used 7.10 Cat or a beta driver ?
3870 CF 20k would be quite nice CF numbers and tells CF performance scales very well with these cards but the price doesn't justify me, 2x239 = $478 vs a single 8800GT for around $250 which should be only around 35% slower than HD 3870 CF. I wonder how CF performance scaling translates into actual gaming performance though, 1.5x~1.9x depending on game I guess.
complete video :
(video) >> http://dl.free.fr/iEhLCf4u4/9CF7C521d01
(Gom Player) >> http://www.download.com/3001-2139_4-...t=dl-GOMMediaP
open video with Gom player
done!!!
:up:
3870 onsale anywhere yet?
Mascaras, the auto clock utility really up your score by 6K points??? I mean that's about impossible...
There here lady's ...
Just bought 2 * 3870 Diamond Viper at NCIX ...
239.99$ cdn each !
Will bench soon for you !
1 year of building my gaming rig ... Arrgh the final moment! ! I can't wait to OC those babies !
What do You mean, the ss I provided You is taken from almost exactly the same frame as your 2nd ss ( actually a frame or two before yours, giving your HD2900XT an advantage as the gentleman I am :D ) ... and there's no notable difference in IQ between our ss.
//
But here it is, the distant ss:
http://i17.tinypic.com/834vlt4.jpg
Take note this time I got EXACTLY the same frame as yours I believe.
Your distant ss:
http://i11.photobucket.com/albums/a1...rysisIQ3-1.png
Spot the huge difference?
:up:
Edit: Come to think of it I do spot differences...the rocks, the truck and bush (to the right edge closer to the bottom, ontop of the cliff) do look a tad more defined in my earlier pic. Another example is a small patch of grass in the bottom portion of the pics (right of center between the 2 boulders and 1 large rock forming a triangle) that look a tad more defined in my early pic. More subtle then anything...
From the review article comparing 3870 and GT, yes I do still see differences. In any case, the LOD excuse has been exercised per my previous post
:up:
Gibberish, what I asked for is cut and dry. Also, as you already proved, the 2 SS are not "almost exactly the same frame" :yepp:
when it's ship ... if all good this week end ...