What do you use to benchmark?

[TrueEevil]

I need to download a benchmarking program and wanted to see what everyone else has.

Was Gonna Download 3dMark05 butim on a 56k connection till next week so the 300mb is a little big (laughs).

 

[Malak]

Most people use them all actually.

 

[Zugzwang152]

I use aquamark3, 3dmark 01, 03, 05. i also use prime95 for stability burn ins.

 

[Insomniak]

Games. Actual, playable software.

What a novel thought.

 

[Zugzwang152]

Originally posted by: Insomniak
Games. Actual, playable software.

What a novel thought.

:roll: benchmarking software has its place and uses. quantifying performance is always useful. in an AT review of a new graphics card you want Anand to say "Yeah, I fired up the game and it looked pretty smooth. but the other card was pretty smooth too. highly recommended!"

 

[Insomniak]

Originally posted by: Zugzwang152
in an AT review of a new graphics card you want Anand to say "Yeah, I fired up the game and it looked pretty smooth. but the other card was pretty smooth too. highly recommended!"

No, I want him to do exactly what he does - benchmark the hardware using realtime game demos.

kthxbye.

 

[Zugzwang152]

kthxbye.

how mature of you. :roll:

 

[Insomniak]

Listen, if you don't want to recieve any flippant responses, something I can understand, then please don't make any ridiculous comments.

Games are the best benchmarks available for measuring total PC performance. If you want to test how the real thing performs, then test the real thing. If you want to know how well photoshop performs on hardware X, then render something and time it.

This is not a complex issue.

Benchmarking software does have its place - mostly as a system stressor. But when it comes to actually benchmarking performance of a given program, I'm going to use that program if it's possible to do so. Anything else is simply decreasing the accuracy of your results, and if you do it deliberately, well, then I have no pity for you.

 

[rise]

wells its alot easier to find comparative system to system scores using well known benchies. obviously it's better to use game demos, but if he's looking for a relatively easy way to see if his sytem is performing decently, then 3dmarks are fine.

 

[gtd2000]

Originally posted by: Insomniak
Listen, if you don't want to recieve any flippant responses, something I can understand, then please don't make any ridiculous comments.

Games are the best benchmarks available for measuring total PC performance. If you want to test how the real thing performs, then test the real thing. If you want to know how well photoshop performs on hardware X, then render something and time it.

This is not a complex issue.

Benchmarking software does have its place - mostly as a system stressor. But when it comes to actually benchmarking performance of a given program, I'm going to use that program if it's possible to do so. Anything else is simply decreasing the accuracy of your results, and if you do it deliberately, well, then I have no pity for you.

I agree with you totally - I've not used benchmarking software since the Madonion days of 3DMark 2001. Even then it was just to see what difference overclocking my graphics card made to the results.

I can feel that overclocking my R9600 in my Emachines M6810 from 296 to 425/200 to 222 makes for better framerates and thats good enough for me.

A dedicated in game stress test is the way to go.

 

[Fern]

Originally posted by: Insomniak

Games are the best benchmarks available for measuring total PC performance. If you want to test how the real thing performs, then test the real thing. If you want to know how well photoshop performs on hardware X, then render something and time it.

This is not a complex issue.

Benchmarking software does have its place - mostly as a system stressor. But when it comes to actually benchmarking performance of a given program, I'm going to use that program if it's possible to do so. Anything else is simply decreasing the accuracy of your results, and if you do it deliberately, well, then I have no pity for you.

Your point is well taken. However, for me at least other benchies do have real value.

After building a rig I like to use benchies to help ensure I have done everything correctly. Not all my rigs are gaming machines, and there's little to no benefit from running actual (office) apps to test for performance.

I also find them useful when upgarding or replacing work rigs. I use various benchies (Harddrive, mem, cpu etc) to measure the performance increase in a useful and quantitative way. Again office apps don't help, or at least the type we use don't.

When upgrading or OCing my gaming rig I like to use gfx benchies to see the increase. But, yes running in-game time demos etc to see increase in FPS is most interesting and relevant.

Fern

 

[Megatomic]

I agree with Insomniak, too. I do use some canned benchmarks though, they are only used to when I make an adjustment to my computer to see if it was detrimental or not. I don't make a practice of buying hardware based on how it does in 3DMark or Aquamark or Sisoft.

 

[Zugzwang152]

I suppose this thread can be chalked up to ambiguity on the part of the OP. The subject line says "what do you use to benchmark?", however in the post, he specifically states that he is looking to download benchmarking apps, which is why I saw Insomniak's post as a complete thread crap and decided to flame him rather than respond intelligently.

 

[rise]

Originally posted by: Megatomic
I agree with Insomniak, too. I do use some canned benchmarks though, they are only used to when I make an adjustment to my computer to see if it was detrimental or not. I don't make a practice of buying hardware based on how it does in 3DMark or Aquamark or Sisoft.

that is what i was trying to say. i use them when i add/change, not as a bragging tool/buyers guide.

 

[Fern]

Originally posted by: Zugzwang152
I suppose this thread can be chalked up to ambiguity on the part of the OP. The subject line says "what do you use to benchmark?", however in the post, he specifically states that he is looking to download benchmarking apps, which is why I saw Insomniak's post as a complete thread crap and decided to flame him rather than respond intelligently.

I wasn't disagreeing with you.

My post was a little "vague". I tried to point out my opinion that benchies DO have their place.

I think b4 we tell somebody to forget benchies and just run games, we need to know WHY they want to run benchies. For all we know, the guy just installed a new CPU and thinks gfx benchies will tell if he set it up right.

In otherwords, our advice should be tailored to HIS purpose. Not ours.

Peace

 

[Insomniak]

Originally posted by: Fern
Your point is well taken. However, for me at least other benchies do have real value.

After building a rig I like to use benchies to help ensure I have done everything correctly. Not all my rigs are gaming machines, and there's little to no benefit from running actual (office) apps to test for performance.

I also find them useful when upgarding or replacing work rigs. I use various benchies (Harddrive, mem, cpu etc) to measure the performance increase in a useful and quantitative way. Again office apps don't help, or at least the type we use don't.

When upgrading or OCing my gaming rig I like to use gfx benchies to see the increase. But, yes running in-game time demos etc to see increase in FPS is most interesting and relevant.

Fern

I'm pretty much in the same boat - keep canned benchmarks installed, but it's mainly to stress test OCs/Burn in upgrades.

 

[DimZiE]

UT2004, FarCry.. i prefer real games

although i also used mark05 and prime95 after overclocking just for testing purpose

 

[HamburgerBoy]

Originally posted by: Insomniak
Games. Actual, playable software.

What a novel thought.

:thumbsup: