1. Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

New 2900XT (new drivers) vs 8800 GTS Review

Discussion in 'Graphics Cards' started by d44ve, Jun 14, 2007.

  1. newtekie1

    newtekie1 Semi-Retired Folder

    Joined:
    Nov 22, 2005
    Messages:
    22,878 (5.86/day)
    Thanks Received:
    8,830
    You know I really don't like how they change the settings in each game for each card. Personally I don't think that is a good way to benchmark and is misleading.

    Instead of keeping all the tests equal and just giving the information about what each card does when presented with the same senerio they change the tests to make the cards performance seem equal.

    I understand that they are trying to give a realistic idea of what the cards are capable of by adjusting the tests to playable real world settings, but at the same time that makes the comparison performance results useless.
     
    10 Year Member at TPU Crunching for Team TPU
  2. bigboi86

    bigboi86 New Member

    Joined:
    Apr 8, 2006
    Messages:
    1,450 (0.39/day)
    Thanks Received:
    35
    Location:
    techPowerUp!
    Go read the review itself guys, they explain why they do the things the way they do.

    If you do a review, at settings above the cards ability, then how would that review be useful to people who want to see how the card performs in real life? Do you think someone is going to play oblivion all laggy? :rolleyes:

    They put both cards at their max ability, then compared them. Sorry if the ATI had to be the one crippled to show decent gameplay, don't blame the [H], blame ATI.

    EDIT: I see where you're coming from newtekie, but that's just how [h] decided to do their reviews. I remember a while back how Kyle(H owner/editor) dissed synthetic benchmarks, and a lot of review sites(including my main forum), saying how their benchmarks are useless. You now see his new methods of reviews. I kind of like it and dislike it at the same time, but I do see his logic in how he does it.
     
    10 Year Member at TPU
  3. DaMulta

    DaMulta My stars went supernova

    Joined:
    Aug 3, 2006
    Messages:
    16,135 (4.42/day)
    Thanks Received:
    1,458
    Location:
    Oklahoma T-Town
    I did read it. I just don't understand why they just don't leave them both on the same setings.

    It's not like the thing black screened when you tried to run it.
     
  4. bigboi86

    bigboi86 New Member

    Joined:
    Apr 8, 2006
    Messages:
    1,450 (0.39/day)
    Thanks Received:
    35
    Location:
    techPowerUp!
    First of all, ATI and Nvidia have different AA techniques(as well as other features). Kyle wanted to show both of them off and compare the both of them, as any reviewer should. Not just spit off framerates.

    Secondly, Kyle doesn't like to review something that wont help the people reading the reviews. If you read this review, you should know EXACTLY which card is better and why. The review does that, doesn't it?

    Would you buy a card that couldn't play oblivion with as high of settings as the other offering? This is the gist of his review.
     
    10 Year Member at TPU
  5. newtekie1

    newtekie1 Semi-Retired Folder

    Joined:
    Nov 22, 2005
    Messages:
    22,878 (5.86/day)
    Thanks Received:
    8,830
    Of course people aren't going to play Oblivion all laggy, but what is the point of doing a performance comparison if you are just going to adjust the settings so all the cards spit out the same performance numbers?
     
    10 Year Member at TPU Crunching for Team TPU
  6. newtekie1

    newtekie1 Semi-Retired Folder

    Joined:
    Nov 22, 2005
    Messages:
    22,878 (5.86/day)
    Thanks Received:
    8,830
    Yes, and that could have just as easily been accomplished by leaving the settings at exactly the same settings and showing that the game is playable on one card and not on the other.

    Now the person actually has to read the review closely to determine which card performed worse and pay attention to the little details such as grass levels and resolutions used for each card.

    If the review was done properly all the reader would have to do is glance at the performance charts read the framerates and they would know the 2900XT tanked in the tests. Instead at first glance [H] made it look like the cards performed equally, and lets face it, most readers don't go past first glances when reading performance comparisons.
     
    10 Year Member at TPU Crunching for Team TPU
  7. HellasVagabond New Member

    Joined:
    Jan 19, 2007
    Messages:
    3,404 (0.98/day)
    Thanks Received:
    162
    Location:
    Athens , GREECE
    Unavoidable results and quite true ( Although i did expect the stock 8800GTS abit slower than this ) if you ask me ;)
     
  8. DaMulta

    DaMulta My stars went supernova

    Joined:
    Aug 3, 2006
    Messages:
    16,135 (4.42/day)
    Thanks Received:
    1,458
    Location:
    Oklahoma T-Town
    I can play that game(x1950XTX) HDR/6AA/16AF with adaptive aa off and I get around 40-50-60 frames at all times on the X1950XTX. Full Grass everything to max at 1280x768.

    I bet the HD 2900XT could do the same.
     
  9. rampage

    Joined:
    May 31, 2007
    Messages:
    1,051 (0.31/day)
    Thanks Received:
    69
    Location:
    Port Fairy, Victoria, Australia
    not bashing the review, it was done well, with alot of effort involved, but i agree with a few ppl here in saysing that the comparisions should ahae done at the same resolution or multiple resolutions for each card, im only sayiing this because i know with my 8800gtx i get a higher frame rate at 1680*1050 then i do at 1280*1024, for example Day of Defeat Source 1280*1024 90>120 fps 1680*1050 110>130 fps.
    so for a fair test each card should be tested at multiple resolutions so you can choose the card that best suites the resolution you run at....
     
  10. Xaser04

    Xaser04

    Joined:
    May 15, 2007
    Messages:
    751 (0.22/day)
    Thanks Received:
    108
    The point of the reviews is not to show which cards can get the highest numbers (which can often be more mis-leading) but to show what are the maximum playable settings each card offers in different games.

    Normally the slowest card will be the benchmark and the quicker ones will have 'gameplay advantages' which basically state what the faster card can offer in this game other the one used as a benchmark )ie 2900xt against a 8800GTS.

    It makes a refreshing change to read a review like this as gives more of a real world test.

    They also make the point that in certain games a high average FPS can be quite mis-leading as the game may actually feel like its lagging despite holding a high fps. (something that would be missed with my line is bigger than your line benchmarks)
     
  11. WarEagleAU

    WarEagleAU Bird of Prey

    Joined:
    Jul 9, 2006
    Messages:
    10,814 (2.94/day)
    Thanks Received:
    547
    Location:
    Gurley, AL
    D44ve, while your way of presenting things are peculiar, I thank you for the link and the review.

    Next time though, try not to take things so hard ::laugh:: j/k
     
    10 Year Member at TPU
  12. d44ve New Member

    Joined:
    Apr 16, 2007
    Messages:
    2,521 (0.74/day)
    Thanks Received:
    309

    Thanks, I will try harder next time :D
     
    WarEagleAU says thanks.
  13. WarEagleAU

    WarEagleAU Bird of Prey

    Joined:
    Jul 9, 2006
    Messages:
    10,814 (2.94/day)
    Thanks Received:
    547
    Location:
    Gurley, AL
    ::laugh:: good deal there buddy.
     
    10 Year Member at TPU
  14. uber_cookie

    uber_cookie

    Joined:
    May 25, 2005
    Messages:
    122 (0.03/day)
    Thanks Received:
    30
    only when full DX10 games come out full potential of 2900 will be seen till then fingers crossed for ATI.
    i love red big noisy demon :D
     
    10 Year Member at TPU
  15. bigboi86

    bigboi86 New Member

    Joined:
    Apr 8, 2006
    Messages:
    1,450 (0.39/day)
    Thanks Received:
    35
    Location:
    techPowerUp!

    My thoughts exactly. Thanks for putting it in better words.
     
    10 Year Member at TPU
  16. EastCoasthandle

    EastCoasthandle New Member

    Joined:
    Apr 21, 2005
    Messages:
    6,889 (1.67/day)
    Thanks Received:
    1,509
    That's not what I'm seeing. What I am seeing is:
    -a pre-retail HD 2900XT being used to show current retail card results
    -pre-release beta drivers being used to show IQ and frame rates

    These are but a few examples of why people have complained about this review. However, other reasons are (but not limited to):
    -differences between the pre release drivers betas and current Cat 7.5 drivers
    -differences between resolutions
    -differences between XP and Vista
    -etc.

    That is only part of the reason why some view the review as bad IMO.
     
    Last edited: Jun 15, 2007
    10 Year Member at TPU
  17. HellasVagabond New Member

    Joined:
    Jan 19, 2007
    Messages:
    3,404 (0.98/day)
    Thanks Received:
    162
    Location:
    Athens , GREECE
    :shadedshu:banghead::laugh:
     
  18. Wile E

    Wile E Power User

    Joined:
    Oct 1, 2006
    Messages:
    24,323 (6.78/day)
    Thanks Received:
    3,777
  19. HellasVagabond New Member

    Joined:
    Jan 19, 2007
    Messages:
    3,404 (0.98/day)
    Thanks Received:
    162
    Location:
    Athens , GREECE
    Different settings , different reviewers, different rigs, different drivers...They all matter.
    However in the last 5 X2900XT reviews ive seen in none does it score 100% better than the 8800GTS. The results are mixed 50/50 and they shouldnt since the 8800GTS has what ? Almost a year on its back while the X2900XT just a month or so.....
     
  20. Wile E

    Wile E Power User

    Joined:
    Oct 1, 2006
    Messages:
    24,323 (6.78/day)
    Thanks Received:
    3,777
    But the 8800 has also had that long to improve their drivers, based on customer feedback. Look back to when it first released, compared to now. They've seen some pretty big improvements.

    The 2900 has been in the wild for only a month.

    And you can't say that ATI had all that time to perfect the drivers. Just like nVidia, they have to rely on customer input. There are just too many possible hardware/software combinations to do all the testing in-house.

    Note, that I'm not saying the 2900 will win in the end. I'm just saying it's still too early to call a victor.
     
  21. EastCoasthandle

    EastCoasthandle New Member

    Joined:
    Apr 21, 2005
    Messages:
    6,889 (1.67/day)
    Thanks Received:
    1,509
    Exactly, if the HD 2900XT was out for 7-8 months and there was no performance increase it would indicate something wrong. But as it stands now it's only been 1 month and 1 driver release making any argument about it's performance moot.
     
    10 Year Member at TPU
  22. Zeratul_uy

    Zeratul_uy New Member

    Joined:
    Apr 28, 2007
    Messages:
    530 (0.16/day)
    Thanks Received:
    38
    Location:
    Montevideo, Uruguay
    Agreed, that way you see if the thing lag and the other thing it's better or not......
     
  23. Zeratul_uy

    Zeratul_uy New Member

    Joined:
    Apr 28, 2007
    Messages:
    530 (0.16/day)
    Thanks Received:
    38
    Location:
    Montevideo, Uruguay
    Wile E you :rockout: and I totally agree with that too :toast:
     
  24. mandelore

    mandelore New Member

    Joined:
    Aug 12, 2006
    Messages:
    3,251 (0.89/day)
    Thanks Received:
    151
    Location:
    UK-small Village in a Valley Near Newcastle
    yeah i run oblivion 1920x1200 pretty much maxed out, withhdr /6aa adaptive on some tweaks and 8xaf and game runs over 40fps outside, and i even have the mod foe extensive weather and foliage etc on my current 1900xtx, the 2900xt is gonna be awsome, so no complaints here
     
  25. HellasVagabond New Member

    Joined:
    Jan 19, 2007
    Messages:
    3,404 (0.98/day)
    Thanks Received:
    162
    Location:
    Athens , GREECE
    Lets not forget that the X2900XT was to be released a few months sooner than it was released in the end due to Driver improvements.....Im not saying that we wont see Bugs fixed in Several Games but i dont think we can expect an big performance boost.
     

Currently Active Users Viewing This Thread: 1 (0 members and 1 guest)