GTX 460 Kombustor Scores

[[TW]Sponge]

^ +1

 

[drunkenmaster]

What tests are consistant these days? I don't really go for benching I just like the ease of afterburner for fan profiles but tried out the test all the same

lots of tests are consistant within 1-2% margin of error, Furmark is NO WHERE NEAR THIS. Its COMPLETELY inconsistant, its so inconsistant I don't think I've ever seen it used in a review with numbers shown, its only used to load the power, which it can't do remotely consistantly. Afterburner is great, kombuster literally does nothing.

@drunkenmaster i know its completely unrealistic but its just a program to see what a card can achieve, and to compare against others, calm down mate!

You can't compare a completely and utterly inconsistant piece of software results against others inconsistant results, with dozens, hundreds, maybe thousands of somewhat accurate benchmarking programs/games/suites out there, why compare with the only one so inconsistant almost no one thinks its worth benchmarking.

The same could be said about Prime95 and Intel's very own LinPack CPU benchmark drunkenmaster, any well designed/built card will not die running code in which it is designed to execute.

The VRM's on motherboards are designed to cope with 100% CPU loads and the same should be the case for graphics cards.

#

No, the same can't be said of that, name me a home use processor that can use well over 350W? Its VERY easy to load up a cpu to 100% power load and it this is the important thing, CPU's and motherboards are designed for use in these situations, if some magical overvolting, ridiculous loud inducing benchmark came out that somehow bumped the rating above and beyond anything else seen, no the motherboards WOULDN'T cope with it.

GPU, GPGPU, nothing, in the past 20 years has ever put a load on like Furmark, its been described by many people as a power virus program, its designed not to be an accurate benchmark, not to be efficient, not to provide a fun experience, not to be a game, its design to do nothing other than put the highest power load possible on a GPU.

Do we all want to have £20 more expensive graphics cards because, for this one piece of software they all spend £20 more on better cooling and higher quality VRM's? Or do we want cheaper cards that will never ever fail or use that much power under load, or stupid cards overbuilt because of one completely useless program? I don't want to pay £20 more to run an inaccurate benchmark, but thats just me.

Keep in mind there wasn't a SINGLE game, or gpgpu type program that would cause a 4870/4870x2 to fail due to power load, theres a super computer with 5000 4870x2's in it, none of them fail under the highest load any 4870x2 in the world goes through.

Furmark could make 4870's crash due to overloaded VRM's.

If you want to compare systems, gpu's, use 3dmark, crysis, something half interesting to watch at the very least, something designed to be a benchmark to compare performance, Kombuster/Furmark is NOT designed for that.

 

[[TW]Sponge]

lots of tests are consistant within 1-2% margin of error, Furmark is NO WHERE NEAR THIS. Its COMPLETELY inconsistant, its so inconsistant I don't think I've ever seen it used in a review with numbers shown, its only used to load the power, which it can't do remotely consistantly. Afterburner is great, kombuster literally does nothing.



You can't compare a completely and utterly inconsistant piece of software results against others inconsistant results, with dozens, hundreds, maybe thousands of somewhat accurate benchmarking programs/games/suites out there, why compare with the only one so inconsistant almost no one thinks its worth benchmarking.

#

No, the same can't be said of that, name me a home use processor that can use well over 350W? Its VERY easy to load up a cpu to 100% power load and it this is the important thing, CPU's and motherboards are designed for use in these situations, if some magical overvolting, ridiculous loud inducing benchmark came out that somehow bumped the rating above and beyond anything else seen, no the motherboards WOULDN'T cope with it.

GPU, GPGPU, nothing, in the past 20 years has ever put a load on like Furmark, its been described by many people as a power virus program, its designed not to be an accurate benchmark, not to be efficient, not to provide a fun experience, not to be a game, its design to do nothing other than put the highest power load possible on a GPU.

Do we all want to have £20 more expensive graphics cards because, for this one piece of software they all spend £20 more on better cooling and higher quality VRM's? Or do we want cheaper cards that will never ever fail or use that much power under load, or stupid cards overbuilt because of one completely useless program? I don't want to pay £20 more to run an inaccurate benchmark, but thats just me.

Keep in mind there wasn't a SINGLE game, or gpgpu type program that would cause a 4870/4870x2 to fail due to power load, theres a super computer with 5000 4870x2's in it, none of them fail under the highest load any 4870x2 in the world goes through.

Furmark could make 4870's crash due to overloaded VRM's.

If you want to compare systems, gpu's, use 3dmark, crysis, something half interesting to watch at the very least, something designed to be a benchmark to compare performance, Kombuster/Furmark is NOT designed for that.

I dont think this guy likes kombustor!