New benchmark suggested

[Benji Tshi]

Hi there

 

I just want to give you my idea.

I don't know how to prevent from cheaters, but it could be a great thing if there is a competition on LCC.

For example, on XS, a lot of guys try Spi32M low clock challenge.

I thought awarding this by cups (no hwboints) could be a good thing.

 

To be continued

[Oskaliber]

Not bad idea. Just cups or for example other points category limited, like hardware points are limited to 300, there could be LCC points limited for 100~ or smth. I think it's worth to consider it.

[chispy]

That idea sounds right , as it will bring something more for the overclocking community , not bad idea at all.

 

 

chispy.

[71proste]

Hi there

 

I just want to give you my idea.

I don't know how to prevent from cheaters, but it could be a great thing if there is a competition on LCC.

For example, on XS, a lot of guys try Spi32M low clock challenge.

I thought awarding this by cups (no hwboints) could be a good thing.

 

To be continued

 

Are you talking about Spi?Like hwbot low clock?

[Benji Tshi]

We can make it on Spi, but there is 3DM01 LCC challenge and of course we can make it with other benchmarks But hte idea is here, the goal is not to make a high score but a real optimized score.

This will give a chance to people wo don't have a lot of money to have some fun (and cups )

[jmax_oc]

Yep, very good idea. I think that Spi32M, 3d01 et maybe Everest Bandwidth should be good

[Benji Tshi]

Everest would be a nice idea too, but due to several version issues, maybe it will be too hard to make a fair ranking...

[3oh6]

with LCC challenges, you have to have a verification method otherwise it is too open to cheating. good faith works over at XS in our thread, but on HWBot? come on. one way is to include in submission rules a digital photo of the benchmark mid stride (loop 20 or something) with CPU-Z open. all that would be required is a photo of the run at loop 20, with the iteration times being visible as well as CPU-Z. it is harder to photoshop a photo...still possible i guess but at least a deterrent.

 

yes having CPU-Z open will hinder the run, but if everyone has to have CPU-Z open for the run, it is still fair.

 

just my $.02

[Benji Tshi]

I don't know how to prevent from cheaters...

 

It was exactly my meaning Jody

But you're true, if everyone have to open a CPU Z window, it's fair enough.

[Benji Tshi]

Read what 3oh6 write

[Benji Tshi]

f*cking cheaters...: you're damned right, unfortunately...What's about everest bandwidth ? too hard to write some rules ?

[71proste]

f*cking cheaters...:

 

exactly mate

the everest bandwidth is good idea too

[Warrior_oF_Byte]

What about FSB ranking?

I think it could be accepted very well, and we can use CPUz like always.

 

CPUz Core speed Benchmark (actual)

CPUz FSB Benchmark (new!)

 

It sounds like, wprime32 and 1024 or superpi 1M and SuperPi 32M.

 

It's just a suggestion!! =)

[K404]

A photo of the monitor mid run would help deter casual cheats, as suggested earlier. The iteration times would have to match the screeny so cheating would be harder.

[Warrior_oF_Byte]

FSB ranking is useless as the architecture of the CPUs changes sometimes in tenfold, without a similar performance boost to go with it. Will be list quad FSB or single? octa FSB numbers? or BCLK ?

 

let's keep things simple

 

 

MMMMMM

 

ok

[noxon]

A video of the run + opening cpuz would do it ?

It isnt hard to do, and the video can be uploaded to youtube or whatever.

[diezel]

fsb benchmark scores

[tiborrr]

This won't work - modern day ASUS ROG motherboards, such as Rampage Extreme, have on-board physical controls to change the FSB on-the-fly. It could be changed after the very first iteration so there wouldn't be visible descrepancies.

 

You just need a buddy on ROG controls while you're filming We would be better off with a repaired validator of SuperPI. Could hwbot programmers fix that?

 

All in all this is a no go.

[tiborrr]

Someone has to has SPi1.4 source, the validator and anti-cheat could be incorporated. 1.4 vs 1.5 results are comparable, so no issue here

[Don_Dan]

I like the idea of adding Everest bandwidth scores, that would be a benchmark totally different from all the benchmark that are now part of hwbot's test suite.

[TheKarmakazi]

why plural?

 

no src is available, so no fix can be made

 

http://www.xtremesystems.org/forums/showpost.php?p=3812090&postcount=19

 

I think validation maybe fixed soon! Problem is only with macos + apache