On the Splinter Cell benchmark

micron

Diamond Viper 550
Veteran
I've read that Dave and Rev have a custom SC benchmark that they use for testing. I was suprised to see another website using it, is it openly distributed? who else has it? http://www.hothardware.com/hh_files/S&V/aiw_9800_pro(7).shtml
In our Splinter Cell benchmarks here, we once again used the handiwork of our friends at Beyond 3D, with their "Oil Rig" Demo test and benchmark script.
Take a look at the screen shot above. See the flat areas in the water? Last we checked, when water ripples from wind or current, large areas don't go completely flat calm at random.
;)
 
Actually, I thought a recent patch added the functionalityto the retail game...? The only thing I thought was unique was earlier access by reviewers like Wavey, and a custom B3D timedemo.
 
No, I made the demo. We've been using it in our recent reviews here. Looks like you don't pay enough attention when you read our reviews :?

http://www.beyond3d.com/downloads/splintercelldemo/

Additional info : New SC demo will be featured in my forthcoming NV34 review. It's based on a different level in the game. Part of our "Preventing App/Demo, er, Optimization" consideration, hehe. Also, custom UT2003 demo in this forthcoming review.
 
Rev, I havent kept up with every benchmark tool, but wasnt there some issues with the SC demo. Something along the lines of water not being rendered properly. Have the issues been correct, since your planning on including the benchmark to your upcoming review?

later,
edit:eek:k, thanks for the link,rev. you beat me to the post. anyways have the issues been corrected with the newer drivers?
 
epicstruggle, see my post just before yours.

The issue doesn't show on 5200 but does on 5600 and higher. NVIDIA's tech guy handling Ubisoft told me :

NVIDIA said:
Hi Anthony,

We've been aware of this bug for several weeks and it has been fixed. Our
next public driver release should contain the fix. The problem was due to a
very subtle behavior difference between the high-end and low-end
architecture.

Best regards,

He knew I was using the 44.03 so we'll just have to see what happens when the next official drivers are out.
 
Thanks for the info, rev. Anychance you can check whether the fix has been placed in the beta drivers? Im sure the new public drivers will be out soon.

later,
 
Reverend said:
BTW, Brent/[H] and us had discovered the anomaly witnessed in HotHardware's 5600 review.
I saw the posts where you and dave were talking about rendering anomolies in SC, I new that HotHardware werent the discoverers of this.
Reverand said:
Looks like you don't pay enough attention when you read our reviews :?
What part of this do you think I missed?
In our Splinter Cell benchmarks here, we once again used the handiwork of our friends at Beyond 3D, with their "Oil Rig" Demo test and benchmark script.
Did I not post this?
*edit*
I knew that in your reviews, you use a custom SC time demo, when I originally posted, I was stating that I was suprised to see that you'd distributed it to other reviewers.
 
Well, we didn't "distribute" it to anyone -- we just made it available to the public for download. The first time this demo was mentioned and used was in one of Dave's reviews (can't remember which one). The link to the download page was in that review and has been in every subsequent review since.
 
Reverend said:
Additional info : New SC demo will be featured in my forthcoming NV34 review. It's based on a different level in the game. Part of our "Preventing App/Demo, er, Optimization" consideration, hehe. Also, custom UT2003 demo in this forthcoming review.

Will those new demos be public?
 
I hope not, there is no need to for us to have them...they should be locked up real good and kept safe from 'optimizing fingers' :!:
 
I think they should use 2 demos one custom private and one custom public and use the private one as a control so to speak.
Or make a set of criteria for creating demos to ensure that demos are generally created using the same type of rendering stressors. Then you can use the review demo once and release it at the time of review. Then create a new one for each review. the only thing that would be questionable would be the strength of the demo creation "rules" and how well a creator stuck to them.

edit
also in regards to the first point maybe a reviewer can just use a private demo only and release it to the public at some arbitrary point in time (i.e. six monthes after creation) at which point the cycle repeats.
 
Brent said:
Reverend said:
Additional info : New SC demo will be featured in my forthcoming NV34 review. It's based on a different level in the game. Part of our "Preventing App/Demo, er, Optimization" consideration, hehe. Also, custom UT2003 demo in this forthcoming review.

Will those new demos be public?
Yes, both will be available as downloads to the public.

I believe in better performance with newer drivers but when the improvements seem unrealistically high, I will likely check timedemo vs actual gameplay performance (using the same level upon which the timedemo was recorded). If actual gameplay do show the performance improvement, I will likely say "Good job" to the drivers team. But if actual gameplay do not show the kind of improvement seen in timedemos, I will mention this to the public and either record new demos based on the same level or on an entirely different level for future reviews and reference what I said.

This is currently the way I am doing things and will likely to be so for the foreseeable future. I will, however, only probably do this sort of thing with every new official driver releases.
 
They should definitely be public, but if ou wanted to wait a few months to release them that would be fine w/ me. If you use secret demos that people do not know what are then you are creating a bad situation.
 
Back
Top