Please clarify the "New Rule"

[Barton]

Does the "New Rule" that allows mods to ignore what are deemed to be "insignificant errors" allow the posting of times from processors in the wrong classifications?

 

For example is it "insignificant" to post up an Athlon XP processor as if it were an Athlon MP?

 

These irregularities are obvious when the processor descriptions in the CPUZ screenies are viewed.

[richba5tard]

No, the rule is meant to be less strict on submissions which have minor errors in the supplied screenshot (eg, no cpu-z memory tab), it has nothing to do with wrong hardware matches.

[Barton]

No, the rule is meant to be less strict on submissions which have minor errors in the supplied screenshot (eg, no cpu-z memory tab), it has nothing to do with wrong hardware matches.

 

I would say that matching the wrong hardware is a signification error

For an example, see the 1500+ Athlon MP class.

 

http://www.hwbot.org/hardware/processor/Athlon%20MP%201500_

 

If I'm not mistaken, according to CPU-World, all of those should be Palomino cores, but some users are posting times for Athlon XP Thoroughbreds in the Athlon MP 1500+ category.

 

Here's an example: See Bwanasoft's #1 Pi Fast time of 82.7 seconds in the 1500+ MP class. That Pi Fast time is shown by his CPUZ screenie to be from an Athlon XP Thoroughbred, not from an Athlon MP processor.

 

http://www.hwbot.org/result.do?resultId=645831

 

Now that's from a popular HWBot member who posts many entries. It's not unreasonable that he might sometimes get one in the wrong class by mistake.

 

However, the entry modification log shows that the result has been checked by a moderator, but the entry is still there.

[Barton]

Ah, I see the badge. Curious, though that will all of Bwanasoft's other benches in that 1500+ category, CPUZ correctly detected his Athlon MP as an "MP".

 

Moreover, all his other screenies show that his Athlon MP is an Athlon MP "Palomino" - as it should be shown.

 

The one and only bench in question is the Pi Fast benchmark. There it shows that the processor is a "Thoroughbred" as well as showing it is an "XP" rather than an "MP".

 

http://www.hwbot.org/result.do?resultId=645831

 

I'm certainly not suggesting that anyone's cheating here. Definitely not that, just saying that it looks in many ways like the wrong processor was posted in error for the Pi Fast time.

 

Check out the CPUZ benchmark screenie for the 1500+ and see how it reports the member's MP as such. (It's in the #3 slot for that bench.)

 

http://www.hwbot.org/hardware/processor/Athlon%20MP%201500_

 

http://www.hwbot.org/result.do?resultId=638891

 

 

Thanks.

[Barton]

Okay I'll respect that and accept your decision. Thanks for looking into it and for replying to the post. :>)

[tiborrr]

It would be nice if we add a rule to include a stepping of your CPU on the screenshot.

 

e.g.: Always add a photo of your CPU's black sticker with stepping, S/N and batch numbers, like I do:

[knopflerbruce]

I don't think it's a good idea. Even if "the gentleman way" would be to tell those who wonder what batch/stepping you've got, it shouldn't be mandatory. It's close to revealing all tweaks in a notepad document:p

 

Plus, it's very time consuming. (make screenshot, take a shot of the CPU, edit it, paste it into the old screenshot).

[tiborrr]

I agree it's time consuming & could easily be faked, but still - it was an idea.

 

Just to let everybody know - the batches on my screenshots are the actual CPUs used in the tests, I swear on my late brother's honor!

[Barton]

I don't think you can do by taking a simple screen shot.

 

First you would need a very good camera to take a very clear photo of the processor itself. Then isolate and enlarge the black label and save that as image. Then use photoshop or some such program to superimpose the black label image onto the screenie of the benchmark run information.

 

For example, our digital camera can't take photos with that degree of clarity. I tried it once to advertise a cpu for sale on eBay. It wasn't clear enough to read the OPN information.

 

The best method seems to be the one we now have.

 

1) Rely primarily on trusting the honesty of members here, and

 

2) Secondarily rely on the CPUZ information - but as jmke says, even that can sometimes be wrong.

[tiborrr]

Okay, here we clearly have Barton mobile (XP-M) chip in a Barton category:

http://www.hwbot.org/result.do?resultId=869604

 

Please, people, sumbit the results into proper category, it's annoying and it eats my golden cups away

 

Best Regards,

N.

[tiborrr]

Please take a look at this:

 

XP-M 2600+ Wprime32: http://hwbot.org/result.do?resultId=869599

XP-M 2500+ Wprime32: http://hwbot.org/result.do?resultId=869604

 

This is why I would like to see steppings, with all the issues this guy had posting in the wrong categories, this sadly looks like a spreading-results-of-one-chip among different chips. With steppings that would prove at least he has the both chips.

 

I really hope I am wrong about him

[knopflerbruce]

Please take a look at this:

 

XP-M 2600+ Wprime32: http://hwbot.org/result.do?resultId=869599

XP-M 2500+ Wprime32: http://hwbot.org/result.do?resultId=869604

 

This is why I would like to see steppings, with all the issues this guy had posting in the wrong categories, this sadly looks like a spreading-results-of-one-chip among different chips. With steppings that would prove at least he has the both chips.

 

I really hope I am wrong about him

 

These two seem legit, there is a difference in the wPrime detection. Which proves it's a different CPU I guess.