What's the deal here

[Mr.Scott]

Question: Are all of these submissions in the wrong category, or is there really a problem with CPU-Z? I've already reported these, and a mod moved them to what is believed to be the proper class, only to have the person who submitted them move them back. Thus, I reported them again. Today I see more of the same type submission from the same person. If it IS a CPU-Z error, how do we really know what type of chip it is? I'd really like to know this, because I have MP's that I'd like to have scored in the XP classes too. BTW, this shouldn't be taken personal by the one submitting them, but his are the only one's that are like that. tia

http://www.hwbot.org/community/submission/1089603_tapakah_cpu_z_athlon_xp_1600_palomino_2259.36_mhz

http://www.hwbot.org/community/submission/1089600_tapakah_pifast_athlon_xp_1600_palomino_63.31_sec

http://www.hwbot.org/community/submission/1089598_tapakah_superpi_athlon_xp_1600_palomino_45sec_375ms

http://www.hwbot.org/community/submission/1089599_tapakah_superpi_32m_athlon_xp_1600_palomino_41min_47sec_718ms

http://www.hwbot.org/community/submission/1089601_tapakah_wprime_32m_athlon_xp_1600_palomino_1min_14sec_789ms

http://www.hwbot.org/community/submission/1089602_tapakah_wprime_1024m_athlon_xp_1600_palomino_40min_12sec_835ms

http://www.hwbot.org/community/submission/1089736_tapakah_wprime_1024m_athlon_xp_1700_palomino_38min_25sec_18ms

http://www.hwbot.org/community/submission/1089733_tapakah_wprime_32m_athlon_xp_1700_palomino_1min_11sec_425ms

http://www.hwbot.org/community/submission/1089729_tapakah_superpi_32m_athlon_xp_1700_palomino_40min_30sec_828ms

http://www.hwbot.org/community/submission/1089728_tapakah_superpi_athlon_xp_1700_palomino_43sec_750ms

http://www.hwbot.org/community/submission/1089730_tapakah_pifast_athlon_xp_1700_palomino_62.41_sec

http://www.hwbot.org/community/submission/1089738_tapakah_cpu_z_athlon_xp_1700_palomino_2382.61_mhz

http://www.hwbot.org/community/submission/1090980_tapakah_cpu_z_athlon_xp_1800_palomino_2382.49_mhz

http://www.hwbot.org/community/submission/1090977_tapakah_pifast_athlon_xp_1800_palomino_62.36_sec

http://www.hwbot.org/community/submission/1090975_tapakah_superpi_athlon_xp_1800_palomino_45sec_797ms

http://www.hwbot.org/community/submission/1090976_tapakah_superpi_32m_athlon_xp_1800_palomino_41min_17sec_188ms

http://www.hwbot.org/community/submission/1090979_tapakah_wprime_1024m_athlon_xp_1800_palomino_38min_39sec_228ms

http://www.hwbot.org/community/submission/1090978_tapakah_wprime_32m_athlon_xp_1800_palomino_1min_7sec_62ms

[Sweet]

Mr. Scott, It is true, are reported, but we are seriously investigating whether these are Athlon MP or XP Cpu.

 

I also see you have added more reports, all under investigation, takes a time.

The crew has the intention of reaching a fair decision

 

TaPaKaH: Obviously we would help send us more pictures of the CPUs used in this benchmark.

 

Regards to both

 

Sw

Edited December 16, 2010 by Sweet

[Mr.Scott]

Thanks Alex. I know it takes time, and you'll try to be fair.

On the pictures though, you could put any picture you wanted up but you'll still never really know if that's the processor that was used, right? So the only real thing you have to go by is CPU-Z, correct? I'm just sayin....

[TaPaKaH]

I provide picture of the actual CPU with the CPU-Z submission usually - these are all genuine XP 1600+, 1700+, 1800+ and not MP.

I guess they are misdetected by the CPU-Z, if you have an XP Palomino I think you can have a go yourself.

[Mr.Scott]

I provide picture of the actual CPU with the CPU-Z submission usually - these are all genuine XP 1600+, 1700+, 1800+ and not MP.

I guess they are misdetected by the CPU-Z, if you have an XP Palomino I think you can have a go yourself.

I'll set up something shortly. I have plenty of Palominos. If you have one set up, why not test ver. 1.55 and see if that does it too. Here's a ver. 1.55, 2000+ Palomino, and that looks fine.

http://www.hwbot.org/community/submission/1073582_benchbruno_cpu_z_athlon_xp_2000_palomino_2093.68_mhz

Here's a ver. 1.56, 2100+ Palomino, and that looks fine too.

http://www.hwbot.org/community/submission/1075684_whitelion_cpu_z_athlon_xp_2100_palomino_1860.4_mhz

I don't know what to say man. Only your submissions are mis-read?

Edited December 16, 2010 by Mr.Scott

[TaPaKaH]

Unless I find a way to fix my NF7 (which has set itself on fire this morning) I won't be able to make any further tests.

I seem not to be the only one with this issue - Blutregen has a misreported 1600+ too (link)

it's funny that on my 1700+ (example) submissions CPU-Z reads the "Athlon XP" CPU string and still thinks it's an MP

[1Day]

I don't know what to say man. Only your submissions are mis-read?

 

Hmm, strange that the first result I looked at in the Athlon XP 1700+ Palomino was misread.

 

 

http://www.hwbot.org/community/submission/875171_tiborrr_cpu_z_athlon_xp_1700_palomino_2136.06_mhz

 

I think there is a systemic failure by CPU-z to correctly read and identify the CPU's in this particular sub-set from AMD consistently.

 

Mr.Scott if you think otherwise please speak plainly and state what it is you suspect.

[Mr.Scott]

Blutregen has a misreported 1600+ too (link)

 

Flagging that one too.

Besides, that was his only submission for that CPU, so it was probably an entry mistake. You had a dozen submissions on three different processors.

Edited December 16, 2010 by Mr.Scott

[Christian Ney]

If it can help, I have got 1600+ 1700+ 1800+ 2000+ Palomino. And 13 different Socket 462 motherboard.

[Mr.Scott]

Hmm, strange that the first result I looked at in the Athlon XP 1700+ Palomino was misread.

 

 

http://www.hwbot.org/community/submission/875171_tiborrr_cpu_z_athlon_xp_1700_palomino_2136.06_mhz

 

I think there is a systemic failure by CPU-z to correctly read and identify the CPU's in this particular sub-set from AMD consistently.

 

Mr.Scott if you think otherwise please speak plainly and state what it is you suspect.

 

Speaking frankly, given that I have shown that all CPU-Z versions work correctly, and that is the only proof that can realistically be excepted, I think the submissions that are supposedly mis-read should be placed in the category that CPU-Z reads them. Any other excuse is simply exploiting a hole in the verification process, and opens the database up to intentional mis-representation of identified processors using the same excuse.

[Sweet]

Well, partners, Cpu-z truly is the most suitable software who guided us in competition on-line, and for everyone, There is no doubt.

 

Also we know that it has some mistakes in the identification of the cpu, few, but it has

Specially with the old AMD and some old Intel.

 

The different tests that they all could do are important to come to a just conclusion

 

If it can help, I have got 1600+ 1700+ 1800+ 2000+ Palomino. And 13 different Socket 462 motherboard

 

Thank you Chris, any help in the topic is very grateful, tell us if they are these AMD, XP or MP and if you prune to do a few tests with the cpu-z

 

Sw

[1Day]

Speaking frankly, given that I have shown that all CPU-Z versions work correctly, and that is the only proof that can realistically be excepted, I think the submissions that are supposedly mis-read should be placed in the category that CPU-Z reads them. Any other excuse is simply exploiting a hole in the verification process, and opens the database up to intentional mis-representation of identified processors using the same excuse.

 

The premise that CPU-z is infallible is sadly not true. I have seen far to many submissions that are perfectly valid submissions but have had some anomaly present in the CPU-z.

 

However I totally agree with you that the current state of affairs is not acceptable and the fluidity in verification that seems to be the case with this particular sub-set of CPU's needs some kind of fix. What that fix is I do not know, as yet. But we are working at it as staff I can assure you.

 

Just so we are one the same page here - Intentional misrepresentation = cheating in my lexicon.

 

Edit: Ah to slow to day I am. Alex you are too quick

Edited December 16, 2010 by 1Day
Sweet beat me to it.

[Turrican]

guys, just look at your l5 bridges on the cpu.

 

if the last one is closed the cpu is identified as an "athlon mp" by cpu-z. if it's open it's recognized as an "athlon xp".

 

some cpus are weird, because on their core they are labled as athlon xp, but the last l5 bridge is closed so they are still recognized as athlon mps.

[Sweet]

Alex you are too quick

 

sometimes, not always

[Christian Ney]

guys, just look at your l5 bridges on the cpu.

 

if the last one is closed the cpu is identified as an "athlon mp" by cpu-z. if it's open it's recognized as an "athlon xp".

 

some cpus are weird, because on their core they are labled as athlon xp, but the last l5 bridge is closed so they are still recognized as athlon mps.

 

Is closed mean that there is a black line on it ?

[Turrican]

here's picture for better unterstanding.

http://www.cpuheat.wz.cz/html/AXP_multiplier/AthlonXP_MP.jpg

 

the "dark" line is the laser cut, so the connection is "open".

[Christian Ney]

here's picture for better unterstanding.

http://www.cpuheat.wz.cz/html/AXP_multiplier/AthlonXP_MP.jpg

 

the "dark" line is the laser cut, so the connection is "open".

 

Ok thanks, so my 3 Palomino with Green PCB are open => XP

And my 2 Palomino with brown PCB are close => MP ?(Damn I thought they were Athlon XP, fortunately I haven't submited them yet.)

What does taht change ? I mean if it's closed or not, it's just to work as multiprocessing system ?

 

(I am not a modder and not a tweaker)

 

Can you just simply update me(by a link or by pm), if there is changing if I will close, open all bridges ?

 

EDIT: I did some search, seems we can change the vcore, cpu multiplier, FSB, cache.....

Is that allowed by HWBot ?

Edited December 16, 2010 by Christian Ney

[Turrican]

you need mps for multi socket a systems.

but even if you "mod" an athlon xp to athlon mp it won't always work in a dual systems for some reason.

 

of course it's allowed to change vcore (wtf?) and the multiplier...

the bridges just control the stock values. of course you have to enter the cpu in the original category no matter if you change the stock fsb\voltage\multiplier etc.

 

btw. plz. test those 3 palominos with the brown pcb. i'm interested what they will show in cpu-z

[TaPaKaH]

my 1600+ = http://www.abload.de/img/dsc047360fby.jpg = open

my 1700+ = http://www.abload.de/img/dsc047378dmx.jpg = open

my 1800+ = http://www.abload.de/img/dsc047387fl2.jpg = open

makes them MP unless I cut it manually and rebench, right?

 

funny that I bought all these Palominos from different ppl and all seem to be "not originals"

[Turrican]

yeah, i think that's the reason why cpu-z thinks those are mp's, although the cpu-core may indicate other.

[Christian Ney]

Which board ?

Chipset I have :

KT133A

741

735

KT400A

KT400

KM133

KLE133

730

KT266

 

And which version of CPUZ ?

Will build the setup atm

[Turrican]

the best would be to use cpu-z 1.56 of course.

plz. try the most important chipsets. (kt133,266,333,400). just boot, start cpu-z and make screenshots.

[Christian Ney]

the best would be to use cpu-z 1.56 of course.

plz. try the most important chipsets. (kt133,266,333,400). just boot, start cpu-z and make screenshots.

 

Ok was trying with the A7V8X. You know what ?

Guess....

I tried with the dead one :o

Will try with the alive one by now.

[Turrican]

lol

 

ok.

[Christian Ney]

Gimme time to install Windows XP, and I will post CPUZ screenshots. (I have to try the AX1800 brown PCB on an another board cause seems it's not working on the Asus )