Benchmarking Thread!

I just noticed that next week's benchmark is Unigen Heaven. I know for a fact that I'm not going to do well on that one. Overclocking doesn't help with tessellation performance on the HD 5xxx cards. :(
 
mad pistol said:
I just noticed that next week's benchmark is Unigen Heaven. I know for a fact that I'm not going to do well on that one. Overclocking doesn't help with tessellation performance on the HD 5xxx cards. :(
Click to expand...

We're doing the HWBot configuration. Ask reflex for the link, I lost it. :p
 
Just to clarify, the heaven install is EXACTLY THE SAME!

The HWbot portion just launches it from your Heaven install location, sets settings automatically, then runs it.

If you already have heaven installed, you do not need to install it again, all you need is the HWbot portion


IF you already have heaven installed, i you can use the "light install" from the Downloads page
 
Last edited:
HWBot is not working on my rig. It changes my desktop resolution to 1600x900 (what the test is apparently being run at -- all monitors clone as would be expected). Heaven boots up, and sits with a black screen showing 5k+ fps in the top right. I can hit escape and bring up the settings, but I can't get it to run. Can we just post whatever settings you want us to run it at and then run it manually? If it's the exact same thing why bother with this launcher program?
 
Sn1PeR said:
HWBot is not working on my rig. It changes my desktop resolution to 1600x900 (what the test is apparently being run at -- all monitors clone as would be expected). Heaven boots up, and sits with a black screen showing 5k+ fps in the top right. I can hit escape and bring up the settings, but I can't get it to run. Can we just post whatever settings you want us to run it at and then run it manually? If it's the exact same thing why bother with this launcher program?
Click to expand...

Ask reflex. I'm tired of him <replace with whining> about the benches. lol.
 
Sn1PeR said:
HWBot is not working on my rig. It changes my desktop resolution to 1600x900 (what the test is apparently being run at -- all monitors clone as would be expected). Heaven boots up, and sits with a black screen showing 5k+ fps in the top right. I can hit escape and bring up the settings, but I can't get it to run. Can we just post whatever settings you want us to run it at and then run it manually? If it's the exact same thing why bother with this launcher program?
Click to expand...

It gives a score.

That is strange that it doesn't work for you.

Are you using multiple monitors? Try maybe disabling extended display
 
reflex99 said:
It gives a score.

That is strange that it doesn't work for you.

Are you using multiple monitors? Try maybe disabling extended display
Click to expand...

Yes I was. I disabled the side monitors completely in CCC and am running a single 1920x1080 monitor now. The program now completely fails :(

hwbotfail2.png

hwbotfail.png


If I set my desktop resolution to 1680x1050 I don't get the failed to resize error, but I do get the same black screen as shown in the screenshot, only it is fullscreen just as if I were running all of my monitors.

When you say it gives you a score, do you mean like this?
heaven1.png

That is just standard heaven (I cropped out all the other settings etc, as they aren't standard yet), but I can match whatever settings you all are using.

Post whatever settings you want this benchmark run at and I can match them with the standard Heaven 2.1, otherwise I don't see this benchmark working for me :( I've spent over an hour tinkering with this so I'm kind of getting frustrated.

edit: I would also vote that we run Normal tessellation, as the Extreme setting was added for the release of the 480 and is therefore pretty biased imo.
 
Last edited:
Sn1PeR said:
Yes I was. I disabled the side monitors completely in CCC and am running a single 1920x1080 monitor now. The program now completely fails :(

http://i766.photobucket.com/albums/xx307/yeSn1PeR/hwbotfail2.png[//IMG]
[IMG]http://i766.photobucket.com/albums/xx307/yeSn1PeR/hwbotfail.png[//IMG]

If I set my desktop resolution to 1680x1050 I don't get the failed to resize error, but I do get the same black screen as shown in the screenshot, only it is fullscreen just as if I were running all of my monitors.

When you say it gives you a score, do you mean like this?
[IMG]http://i766.photobucket.com/albums/xx307/yeSn1PeR/heaven1.png[//IMG]
That is just standard heaven (I cropped out all the other settings etc, as they aren't standard yet), but I can match whatever settings you all are using.

Post whatever settings you want this benchmark run at and I can match them with the standard Heaven 2.1, otherwise I don't see this benchmark working for me :( I've spent over an hour tinkering with this so I'm kind of getting frustrated.

edit: I would also vote that we run Normal tessellation, as the Extreme setting was added for the release of the 480 and is therefore pretty biased imo.[/quote]

1. It gives a score like 3dmark does. Example: [url]http://hwbot.org/community/submission/2113022_reflex99_unigine_heaven___basic_preset_dx9_radeon_hd_3850_720.8_dx9_marks[/url]

2. How is it biased? If it was optimized to run better on nv cards, then that would be biased, but just using a feature of DX11 that AMD cards suck at is not really biased in anyway.

3. if more people have a problem similar to yours, we will consider changing to the regular bench.

(btw, i have a 6870, so, not really a nv "fanyboy" as they call them)
Click to expand...
 
reflex99 said:
1. It gives a score like 3dmark does. Example: http://hwbot.org/community/submissi...sic_preset_dx9_radeon_hd_3850_720.8_dx9_marks

2. How is it biased? If it was optimized to run better on nv cards, then that would be biased, but just using a feature of DX11 that AMD cards suck at is not really biased in anyway.

3. if more people have a problem similar to yours, we will consider changing to the regular bench.

(btw, i have a 6870, so, not really a nv "fanyboy" as they call them)
Click to expand...

I stated that it is biased in my opinion because it was a feature they added after the fact, at the launch of the 480, in order to show a difference between the 480 and the 5870. I honestly don't care either way, I am just stating that the "update" of a benchmark to add a new feature "extreme tessellation" to coincide -- literally the week of the release of the flagship nvidia card at the time is clearly bias. The normal tessellation provides more than what you see in any game on the market right now anyway. It seems like the HWBot guys have a workaround and are reducing the tessellation quality in their respective gpu control panels to in a sense cheat the system. Regardless, I don't care either way and am not trying to be a fanboy here either. I'm a nvidia shareholder, came to this card of 2x 8800GT's and bought a 460 for my brother last month.

What I do care about is being able to participate in this benchmark and it seems unnecessary to tack on a 3rd party front end simply to get scores like that when the vanilla benchmark provides what I would consider adequate scoring.

see:
Unigine2011-02-1419-36-53-33.jpg


Here is hardocp's article on the original Heaven release:
http://www.hardocp.com/article/2009/11/06/unigine_heaven_benchmark_dx11_tessellation

Some articles on 2.1:
http://www.ngohq.com/news/17897-unigine-heaven-benchmark-2-1-a.html
http://www.hardocp.com/news/2010/03/24/unigine_heaven_20_changes
 
Last edited:
We are probably going to do extreme tess no matter what..

Screw it, bass, what do you say? How bout we go back to normal heaven.

Settings like this:
DX11
high shaders
extreme tess
16x AF (anisotropy)
8xAA
1680x1050
 
We're doing normal, screw it.

Great job, reflex. You just made HWBot look totally fail. :D
I'll have the settings up in a sec.
 
Ph?bus said:
So does the normal Unigine. o.O
Click to expand...

But i cannot submit normal heaven for boints

<--struggling boint farmer

Ph?bus said:
We're doing normal, screw it.

Great job, reflex. You just made HWBot look totally fail. :D
I'll have the settings up in a sec.
Click to expand...

Settings like this:
DX11
high shaders
extreme tess
16x AF (anisotropy)
8xAA
1680x1050
 
Can we do normal tess instead? Make it slightly more fair?

Also, those with non-DX11 cards that can only run this in DX10 will be owning most of us.
 
You must log in or register to reply here.

Latest posts

Back
Top