Far Cry 61.76 Vs 61.77

[Dot50Cal]

No performance gains were noticed except for a fluke on research in 1024 with 4xAA and 8xAA.

http://img70.exs.cx/img70/9743/1024aa.gif

http://img70.exs.cx/img70/3/1024noaa.gif

http://img70.exs.cx/img70/4675/1600aa.gif

http://img70.exs.cx/img70/6986/1600noaa.gif

Image quality comparisons show the same levels.
61.77:
http://img70.exs.cx/img70/3171/new1.png
61.76:
http://img70.exs.cx/img70/7101/old1.png

Notes:
-Sound was ENABLED

-Trilinear was forced via the control panel with all optimizations set to off.
The system that the tests were run on is as follows:

Pentium 4 2.4c @ 3.0
Geforce 6800 GT @ 400/1100 (Default is 350/1000)
1GB OCZ PC4000 3-4-4-8 1:1
400 Watt Antec PSU with 18A on 12+
Abit IC7-G 875p motherboard

[Sazar]

did you retest that research level?

to see if the test indeed was a fluke?

or did you run through it one time?

[Dot50Cal]

Upon retesting with the 61.77's I got 68 FPS. However taking into account the 1600x1200 performance and the fact that Far Cry's demos have a tendancy to not render the same exact things (IE people can be several steps ahead of where they should be) I can only beleive its a fluke. I really dont want to re bench with the 61.76's as Im running the 61.77's right now. If there were a performance increase we should have seen them in the 1600x1200 AA tests as well.

[Sazar]

k.. I was just wondering if you had indeed rebenched it :)

was just clarifying...

nice job btw...

[betterdan]

Ok I did a few benches and saw really no change from 61.76 no sm3.0 and 61.77 no sm3.0 or with both drivers and sm3.0 enabled. There are some instances where 3.0 is making a difference however as i just started up farcry and took these screenshots with 3.0 enabled and disabled. These shots were at 1280x1024 everything maxed including water at ultra high, 4xAA and 4xAF. Trilinear optimizations on and AF optimizations off in the control panel. Using 61.77 and final directx 9.0c SP1 (no beta SP for me :))

Far cry SM2.0 76FPS (http://www.zippyimages.com/files/86048/FarCry1_2.0.jpg)
Far cry SM3.0 91 FPS (http://www.zippyimages.com/files/86049/FarCry1_3.0.jpg)
Difference of 15 FPS at same place in map

Far cry SM2.0 98 FPS (http://www.zippyimages.com/files/86050/FarCry2_2.0.jpg)
Far cry SM3.0 112 FPS (http://www.zippyimages.com/files/86051/FarCry2_3.0.jpg)
Difference of 14 FPS at same place in map


Damn it looks like you will have to right click the link, copy it and paste it into a new browser to see the shots.

[betterdan]

Removed the info cause I was a dumbass :retard:

[Dot50Cal]

Just found this out...
You don't need to start farcry with any devmode or r_sm30path 1 commands anymore. If you have directx 9.0c, the 61.77 drivers, the newest fxc.exe file and farcry 1.2 just start it up and it's automatically enabled without the crappy yellow words at the top. To see if it's enabled just bring down the console and type \r_3m30path then hit enter. It should say \r_sm30path=1 which means it is enabled.

Argh! Its not true :)
"D:\Crytek\Far Cry\Bin32\FarCry.exe" "r_sm30path 1" works without dev mode. The console does say its on, however it isnt. If you type \r_sm30path 1 in the console you're FPS will go higher (certain areas).

[betterdan]

Sorry I got a little excited and posted it in a few place hehe
Ok you were right :) I'm gonna delete my false info :inform:

[Dot50Cal]

NP, You can just edit it until a mod comes along and deleted it.

[betterdan]

Mods please delete my last 2 posts

Ok I did some benches with the ubisoft timedemos and got the following.
All tests were run at Resolution: 1280×1024
Maximum quality option, Direct3D renderer
Antialising: None
Anisotrophic filtering: 1×
no sound
Using only 61.77 drivers

Regulator
SM2.0 Score = 49.19 FPS
SM2.0b Score = 49.83 FPS
SM3.0 Score = 50.93 FPS

Research
SM2.0 Score = 74.74 FPS
SM2.0b Score = 74.62 FPS
SM3.0 Score = 83.30 FPS

Training
SM2.0 Score = 54.42 FPS
SM2.0b Score = 53.97 FPS
SM3.0 Score = 53.77 FPS

Volcano
SM2.0 Score = 76.93 FPS
SM2.0b Score = 76.67 FPS
SM3.0 Score = 83.46 FPS

[anzak]

Mods please delete my last 2 posts

Ok I did some benches with the ubisoft timedemos and got the following.
All tests were run at Resolution: 1280×1024
Maximum quality option, Direct3D renderer
Antialising: None
Anisotrophic filtering: 1×
no sound
Using only 61.77 drivers

Regulator
SM2.0b Score = 49.83 FPS
SM3.0 Score = 50.93 FPS

Research
SM2.0b Score = 74.62 FPS
SM3.0 Score = 83.30 FPS

Training
SM2.0b Score = 53.97 FPS
SM3.0 Score = 53.77 FPS

Volcano
SM2.0b Score = 76.67 FPS
SM3.0 Score = 83.46 FPS

Uhh since when is 2.0b enabled? So much happend while I was gone today.

[Dot50Cal]

Uhh since when is 2.0b enabled? So much happend while I was gone today.

I beleive the command is \r_sm20bpath 1. I dont know if it runs correctly on the 6800's though.

[betterdan]

Well I was using the HardwareOC Bench Tool and in it I selected to enable SM2.0b and it ran fine. maybe i should have selected default model instead. Hmmm

Edit: ok I updated the scores with sm2.0. I don't know if my card was really running sm2.0b but that's what I had the HardwareOC BenchTool set to.

[ststephen]

I got a 200 point drop in 3dMark and small drops in Far Cry benchmarks:

Far Cry:
Ver,SM,Map: Run 1/Run2
6176,2,Training: 55.17/53.93
6176,3,Training: 51.89/54.06
6177,2,Training: 54.04/53.25
6177,3,Training: 51.46/53.96
6220,2,Training: 51.75/50.89
6220,3,Training: 49.43/52.15
6176,2,Research: 59.45/60.07
6176,3,Research: 68.55/78.68
6177,2,Research: 58.10/58.71
6177,3,Research: 64.58/76.36
6220,2,Research: 52.70/53.57
6220,3,Research: 60.82/69.99

3DMark03:
6176: 11963
6177: 11745

P43.2@3.52 (220FSB)
1GB DDR 4000
GeForce 6800GT, Audigy2

/// No good for me, I'm switching back.

[Dot50Cal]

61.76 - 11970
61.77 - 11987

:ORDER: You should bench again, I first tried it and I got 10000! I knew something was wrong so I restarted and it solved it. I actually gained a few points.

[ststephen]

You're right. I boot into Linux and use PartImage to restore a clean XP+SP2RC2 image, then install infinst.exe, then nvidia driver, then audigy drivers. The previously saved image was getting ~11700 repeatedly -- three times, across mulitple reboots. Then I did a fresh install again and got 11950.

Bizarre, but comforting, I guess.

[ststephen]

No one cares, but I'd been running my memory at auto, so the P4C800-E was choosing 3:2. I set it to 1:1 properly, now I get 12300 3dMarks joy.

Now if I could only get the memory to run faster. I really messed up when I bought this 3.2, I can only O/C it to 220, so my memory doesn't run that fast even though I have PC4000 memory.

[Blacklash]

Lately I have been trying my rig at 3.68 with 2-2-2-5 timings on the ram and my mobo's GAT at Street racer, which is one up from what I can run at 3.9. The difference with the 61.77 driver(3dmark03) is now 13241 at 3.9 and 13198 at 3.68. The tighter ram timings and higher acceleration seem to offset the higher clock to a degree. We are only talking about 43 3dmark03 marks.Another thing I like is I can run my CPU at default voltage for the 3.68 instead of the mandatory 1.675 to get 3.9.

The 61.77 driver does great with Far Cry. Anyone running everything at very-high/ultra high that wants to see a real FPS boost at 1600 res, just drop your shadows to medium. Yes you lose the nice shadows on your weapon but you gain a large amount of FPS. One point in training for me went from 63 FPS all the way up to a steady 80.

[ststephen]

I'm up to 12475 3dMarks. I can't get my PNY to O/C past 405/1100. Malificar: what speed is your GT at?

[Blacklash]

I'm up to 12475 3dMarks. I can't get my PNY to O/C past 405/1100. Malificar: what speed is your GT at?

It will only do 415/1.20 tops. Sometimes the test fails on 1.20. I wish I could get more on the core. Those Ultra folks getting 460 or so on the core make me sick :D

Edit: I tried my rig at 3.74(220) 2-2-2-5 and I got 13,219 in 3dmark03, so that's what? About 22marks or so over 3.68? I will likely just stay here. Nice to see it can hold the timings there though.

[ststephen]

Yeah but trust me as a recent owner of a 9800XT, a $550 (w/ tax) video card, you feel like a total chump having paid $500+ for a video card when it no longer is the tastiest in 2 months. I feel very happy with my $400 video card -- I don't feel like as much of a sucker. Plus I unloaded the 9800XT for $120, which helped mitigate the cost even more.

I think I did really good with this generation by sticking with the GT. I felt too much like a crack-addled loser paying $500 for a video card that's only cherry for 1 gaming cycle of a few months.

[Blacklash]

Yeah but trust me as a recent owner of a 9800XT, a $550 (w/ tax) video card, you feel like a total chump having paid $500+ for a video card when it no longer is the tastiest in 2 months. I feel very happy with my $400 video card -- I don't feel like as much of a sucker. Plus I unloaded the 9800XT for $120, which helped mitigate the cost even more.

I think I did really good with this generation by sticking with the GT. I felt too much like a crack-addled loser paying $500 for a video card that's only cherry for 1 gaming cycle of a few months.

I agree. This one is even all the more sweet because BB honored the brief 299usd deal they had up for it. I planned on trying a X800XT myself, but I couldn't turn down a card like this for that price. Overclocked it is very close to what I would be getting out of the XT.

This generation it seems like you can get much more bang for your buck. Even the 6800, mildly overclocked it a very good deal for 299 msrp. I have seen them do up to 10k in 3dmark03 depending on the OC you get.

Too bad Ati hasn't got the lower clocked version of XT out the door yet.

Sounds like you are set though, a rig that can do 12.4k in 3dmark03 will handle anything coming out.