S.T.A.L.K.E.R. Performance

NVIDIA has been pushing this game since the GeForce FX days, and now it is finally here. We have to use FRAPS to benchmark it even though a demo_play feature is included. Apparently the demo_record facility had to be removed at the last minute.

S.T.A.L.K.E.R.


Interestingly, at 1024x768, both the 8600 GTS and 8600 GT perform well compared to their competitors. The 8600 GTS manages to beat the X1900 XT 256MB while the 8600 GT keeps up with the X1950 Pro. Unfortunately, the new hardware can't keep its lead beyond 1024x768, as moving up to 1280x1024 gives us a much sharper drop on G84 than on other hardware.

With the added texture address capability per shader and drastically reduced memory bandwidth compared to other 8 series hardware, the problems we are seeing could be related to memory pressure. This would also help to explain the poorer than expected performance in older games that rely more heavily on lots of textures rather than shader speed. We would need to do many more synthetic tests to really get to the bottom of this issue though.

At 1280x1024, a resolution many midrange gamers run, the new 8600 series just isn't able to handle S.T.A.L.K.E.R. with all the settings turned up except grass shadows. Some settings will need to be lowered a bit to achieve better performance at higher than 1024x768.

F.E.A.R. Performance Supreme Commander Performance
Comments Locked

41 Comments

View All Comments

  • nullpointerus - Tuesday, April 24, 2007 - link

    quote:

    Heres to hoping AMD/ATI and nvidia both have great ~$130 DX10 cards later this year...

    Methinks this would be a good time for a cynical retort, but I can't think of one ATM.
  • coldpower27 - Tuesday, April 24, 2007 - link

    We'll see, if the RV630 is competitive to these cards as it too has a 128Bit pCB as well from the rumors we have been hearing.
  • yacoub - Tuesday, April 24, 2007 - link

    Thank you for confirming the mediocrity of these cards. It is good to see a nice unbiased review that does not sugar-coat the pathetic results these cards are posting in some games and the so-so to just-okay results in other games.

    The fact remains a new generation of cards should outperform the previous generation at the same price/performance level.

    The 8600GT needs to CONSISTENTLY outperform the 7600GT, the 8600GTS needs to CONSISTENTLY outperform the 7900GS/7900GT, and neither of these cards can manage that because they are cut down too far. A 256-bit bus and 64 stream processsors on the GTS for the same price point and it would have been a success and ensured NVidia victory across the board.
  • coldpower27 - Tuesday, April 24, 2007 - link

    Which would basically nullify any cost advantage to production the 8600 Line has, if you add a 256Bit PCB plus the increase to 64 SP, you have a card that costs more then the 7900 line to make. There wouldn't be any point, Nvidia would have to eat the costs and still sell at the mainstream price points and make less money in the end.
  • yacoub - Tuesday, April 24, 2007 - link

    7900 series is going away, my friend.
  • CalvinHobbes - Tuesday, April 24, 2007 - link

    I'd love to see some video benchmarks using CPU's that the users of these cards would most likely have. If I were building a system with an 8600 based card I would probably match it with an E4300, E4400 or 5000+.
  • ssidbroadcast - Tuesday, April 24, 2007 - link

    I got your back. If you examine hardware surveys, IE Steam uses over a million unique samples, most people aren't using top-dollar processors or 2GB corsair RAM (pricey stuff). I realize the obvious benefits of using this kindof hardware for benching; cross-standardization being one, but most of us can't really afford these kind of systems.

    It'd be nice to see "modest" benchmarks on an entry-level AM2 machine, or even better, a 939 socket? Pentium equivalents as well?
  • A5 - Tuesday, April 24, 2007 - link

    They're testing the card, not the system. If they used it with a slower CPU, it would skew the results of the better cards a little lower and make the 8600s look better than they are.
  • Sunrise089 - Tuesday, April 24, 2007 - link

    The number one reason nVidia has had more success than ATI in the past few generations is their superior midrange cards. (The number 2 reason is product delays with new high-end part intorductions, but that isn't as severe, since the real high-end market segment will just buy the new cards anyways). nVidia has clearly dropped the ball here, and these new benches confirm what the last review seemed to indicate - that unless AMD is totally asleep, they can nail nVidia with the Radeon 2600 line.

    "The problem is that there is a huge performance gap between the 8600 GTS and the 8800 GTS 320MB." - This is the essense of the problem...the much better card is just priced much to close to these parts. The overall bottom line might have actually been better if nVidia hadn't released the 320meg part at all.

    "We also have multiple cases where NVIDIA's new offerings perform lower than similarly priced hardware from their own previous generation hardware." - This is when it becomes outright embarrassing. It's been a long time, if there ever was a time, when this was true for the boys in green.

    I'll say it again - 64 shader part at $200 will come quickly if AMDs competing parts are any good.
  • SalientKing - Tuesday, April 24, 2007 - link

    Ive been wondering how 2 8600GT would fair

Log in

Don't have an account? Sign up now