Real HL2 benchmarks
- Thread starter imported_obsidian
- Start date
why did they use 6800GT OC and 6800ultra OC? If you ask me, this is more a test showing BFG performance as opposed to 6800 series performance. Also they are using a newer nVidia driver, which was no doubt released just in time for HL2. Good to see the nVidia perfomance gains although I don't see why these necessarily contradict DH since the same conditions were not applied.
They probably did because those were the cards they had. The overclocks on both are fairly minimal.
RussianSensation
Elite Member
- Sep 5, 2003
- 19,458
- 765
- 126
Tom's Hardware is supposed to post their benchmarks of HL2 as well, so that is something to look forward too shortly.
wow, kind of surprising in terms of how well the nvidia cards did, i'm glad we don't have to buy seperate cards for all of our favorite games.
No surprises here; the benches look comparable to the Source Stress Test benches that were conducted at reputable sites like AT. DH's benches (that started all that fuss yesterday) looked completely contrary and it shows now...it's safe to say that the XTPE is marginally better (read: almost unnoticably) than the Ultra in HL2, but it's doesn't take it's lunch money like DH wanted us to believe. All the cards are good for HL2...which anyone with a GT could have told us yesterday...I personally don't care bout the benches since I was too busy playing HL2 with ultra smooth framerates all yesterday =)
these show the x800pro winning (over GT) every time under the most demanding circumstances and they don't even test an ultra. and drivers?
What I don't get is... how are DH fudging the numbers?
yea, but how are they getting such poor results. Are they running Prime95 while testing the 6800GT? Are they blow drying the GT GPU while its running? Are they shaking the computer during the benchmark? Or are they simply making up numbers without even running the benches?
It's just preposterous to imagine any of these scenarios. But maybe its true...
FYI, +r_fastzreject 1 is disabled for ATi cards, and enabled for NV cards in the current state of the game. It should up on for both cards in a future update. The command enables Z fill geometry pass in order to speed up rendering of the full scene rendering pass.
Yeah, DH's numbers were a little strange. I have HL2 and my 600GT overclocked to ULTRA smokes through this game like a knife through butter, even with AA and AF turned way up. If you have an NVIDIA card I think you are set playing this game and any other game in the near future. Both ATI and NVIDIA have really some great products this cycle....
Tho the BFG OC series doesn't add much speed, it's still curious Brent didn't once note their clocks. (OTOH, are the XTPEs even in retail stores anymore?)
His conclusion is reasonable, tho. Even a 10% difference is pretty much inconsequential, especially at 100fps.
His conclusion is reasonable, tho. Even a 10% difference is pretty much inconsequential, especially at 100fps.
Good point Pete, if you go to NE they don't even have a search option for XT-PE anymore, just XT...but when you look around at the XT's you see most at 500 core but some at 520 core (but aren't PE), and the few that are listed as Platinum at 520 as well. It seems like they stopped calling them them PE's and let the individual manufacturers take the risk of that extra 20 mhz on a individual basis...probably couldn't bin them at those 520 safely, which makes sense since the XT/XT-PE is now the last card from the May x800/6800 paper launches to hit the market. Either way, the 20 mhz between XT and PE and the 25 between BFG OC and normal Ultras seems to even it out, and though it is questionable for Brent to do that, they are all at marketed speeds so I guess that makes it slightly more reasonable.
Originally posted by: JackBurton
Man this is great, my 6800GT plays Doom3 AND HL2 great!!! <ME> Hugs his 6800GT.
best card to buy in my opinion... runs everything flawlessly. im running hl2 full detail, 4x/8x @ 1024x7... and i havnt felt any lag yet, at ALL. i love it
Elfear
Diamond Member
- May 30, 2004
- 7,169
- 829
- 126
Well, in my humble opinion, most guys willing to pay $400+ for a graphics card aren't going to play games at 1024x768 and no AA or AF. A lot of the benchies around the net show the X800 series cards having a substantial gain over the 6800 series cards at 1600x1200 with 4AA/8AF which is where a lot of gamers I know like to play if they aren't limited by either their monitor or their video card.
I will agree with everyone that the 6800GT is a very good card for the money and will play any game at most any setting you want. At Doom3 the 6800 series cards simply own the current crop from ATI so really it depends on which games you like, where you like your settings at, and what you can get your hands on at the time. At 90% of all possible settings I don't think you'd be able to tell the difference between the two company's cards.
I will agree with everyone that the 6800GT is a very good card for the money and will play any game at most any setting you want. At Doom3 the 6800 series cards simply own the current crop from ATI so really it depends on which games you like, where you like your settings at, and what you can get your hands on at the time. At 90% of all possible settings I don't think you'd be able to tell the difference between the two company's cards.
RussianSensation
Elite Member
- Sep 5, 2003
- 19,458
- 765
- 126
Firingsquad Review
Here is another review. ATI cards seem to do much better in the shader intensive (water) level.
Also, Half-Life 2 CPU Shootout: AMD versus Intel
Here is another review. ATI cards seem to do much better in the shader intensive (water) level.
Also, Half-Life 2 CPU Shootout: AMD versus Intel
i'd like to point out that all of you are neglecting this part: http://www.hardocp.com/image.h...lnbngzY3hfMl8zX2wuZ2lm and all of the other parts like it. these are run at max quality, max res, and show the x800xtpe more than doubling the 6800Ultra's scores.
that's right, DOUBLE, that's even more than driverheaven said it was...
that's right, DOUBLE, that's even more than driverheaven said it was...
I think everyone here is aware that firing squad and a couple of others got results similar to DH, but don't really like to admit they were wrong. Or worse Anands benches weren't representative.
They're not running at the same settings, the XT PE is running 6XAA (multri sampling) while the Ultra's running 8sXAA which is SuperSampling that renders the scene at a super high resolution then scales it back to the regular resulotion which looks WAY better than multi sampling and consumes WAY more graphics power.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 24K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
-
Facebook
Twitter