The official HWBOT Team Cup 2015 - SC1: CPU Challenge thread.

[Johan45]

That may be so websmile, I have heard that some of the older HW was problematic but the validations that were attempted with new HW that are rejected are the result of what I posted earlier. Some boards the HTT frequency will fluctuate and when "I" was trying to validate on the lowest reaches of that fluctuation it was rejected. When I validated at or near the actual BIOS setting it was accepted. I guess my point was, HWBot already has a specific set of rules that we follow daily. My interpretation is that the rejected results are not valid so by the rules should be disallowed. If that CPU/board won't validate where you want it to then pick something else. I just "assumed" the spirit of the competition was to "find" hardware that would give you an accepted validation at the lowest speed you could get.

[websmile]

Yes, it will be hard to seperate this. If we follow the rules completely, we will have to remove lots of scores, also new hardware scores where fse screenshots show no subtests for example, in spite of validation link , results wit mem tabs showing no freq and so on. I wonder if any results will be left in the end ^^

[Johan45]

Sure would be interesting. I realize this is a big job and decisions made will never make everyone happy.

[Marquzz]

Yes, please no low clock / slowest again. I love overclocking old/slow stuff, so I don't have anything against that, but slowing things down goes against the whole concept, imho.

[Johan45]

I've had fun with it to be honest. It was a nice change from the everyday pushing and had challenges of it's own.

[xxbassplayerxx]

I've had fun with it to be honest. It was a nice change from the everyday pushing and had challenges of it's own.

 

Sounds like someone who didn't have a 20 day run fail

[trodas]

websmile -

results wit mem tabs showing no freq and so on

 

On many chipsets CPU-Z did not report memory frequency and even memory timings. Good example is SIS 730 (PCchips M810LR) or i430TX (Asus TXP4-X). Nothing can be done with it.

 

On the other hand, new CPU-Z version entierly f*cked up timings detection on VIA PT800 chipsets (ASRock 775Dual-VSTA) and the results are completely wrong. This is 3-3-3-10 (28) timings on Crucial Ballistic, thx to Jazzman:

 

 

So you see - by CPU-Z is: CL - unknown, tRAS 9, when it is 10 and no info on tRC, despite older CPU-Z versions did not give any trouble on this mainboard. And I reporting bugs to Frank left and right, but no fix...

 

In case of i430TX I even send him the Intel docs a pointed out what bites on witch registers one can detect the memory settings... what more can I do?

 

...

 

Other programs have quirks too... sometimes it is not that hard to do some record, as to submit the validation. This is, "thanks" to endless bugses of almost everything (for example Aquamark wrapper - run on nVidia Vanta, get score 0, save as valid... submit... and it is still valid ) ...

 

And it is not like my Vanta is bad, it can do 3D well:

...even with slight overclock from 100/125 to 120/160 But the wrapper suxx...

 

Same for Heaven wrapper. Also does not work for me on some machines (and much less weird machines that, this MSI K7TM Pro is)...

 

So, it is fighting with bugses all around the way!

[websmile]

I know this, and people who ask for removal of scores because of Hwbot rules should know this as well - this is the point, you cannot apply rules of verification for hardwae that is 15 years old and that noone has tools for anymore to verify according to new rules. I opted to ask CN to let a specialist for old hardware do the moderating because either way it will be done, people will complain and I haven´t enough time atm to ban them all

[Johan45]

Sounds like someone who didn't have a 20 day run fail

 

Good call, it hasn't failed yet and like Rasparthe I live in the country so power can be ifffy. I've just been lucky, now if it can hold out for another day or so.

[xxbassplayerxx]

Good luck!

 

 

 

 

 

 

 

 

 

 

*Fail fail fail fail*

[trodas]

websmile -

I know this, and people who ask for removal of scores because of Hwbot rules should know this as well - this is the point, you cannot apply rules of verification for hardwae that is 15 years old and that noone has tools for anymore to verify according to new rules. I opted to ask CN to let a specialist for old hardware do the moderating because either way it will be done, people will complain and I haven´t enough time atm to ban them all

 

Yea, the verification problem is awfull. Back in the day, CPU-Z worked pretty nicely and I can verify 10.7MHz, altrough it did not detected the FSB:

http://valid.x86.fr/9b88hu

 

However todays the CPU-Z fail miserably to detect CPU clock and report false FSB clock 50MHz:

 

 

At least it from THAT did not produce "valid" result. However it can produce crazy "valid" results none the less. For example this one:

http://valid.x86.fr/b0zd5t

 

384MB of ram, consisting of 128 + 512MB sticks, detected as 128 + 256MB it report as 640MB of ram. Is not that funny? Definitively CPU-Z shoud allow some moderation in form that user can change the wrongly detected infos under some human supervision, of course...

 

Because I have "valid" result with 640MB ram on i430TX, that is bound to support only 256MB of ram :nana: :nana: :nana:

 

And no validation of 10.7MHz. Oh, well. At least ATI Rage XL give a nice image at this superslow clock

[fgi]

Ops

Edited September 18, 2015 by fgi

[saint19]

I can't submit score in stage 7.

Error is "Validation error: A valid futuremark compare url is required if want to reach the PCMark7 Hall Of fame. As only 1 benchmark scores are better than you, you need to provide additional verification."

I have valid url.

 

Try using https...

[fgi]

very fast reply

10x

[dejo1]

is samsungs "rapid" going to be allowed in the pcmark7 stage?

[ScunnyUK]

is samsungs "rapid" going to be allowed in the pcmark7 stage?

 

 

Id like to know this too but not just Samsung version, is any ram cache / ram drive allowed as I thought using them is illegal?

 

Also I think I know the answer but conformation would be good on whever or not Skylake systems can be used in any stage?

[viper-rd]

Dear Massman.

Explain please - for some reason, my result has been removed from the closed stage "CPU CHALLENGE - HM8X SUPERPI 32M"?

If it is done on a complaint TASOS - then why did only away my result?

[viper-rd]

Only the first three results are correct?

[websmile]

I am no expert, but your result was checked and chipset was not H86/87 but VIU, on result of the lenovo durian that was now reported several times, this is hm86 http://hwbot.org/hardware/motherboard/durian_7a1/

Hope I could help

 

Edit - no skylake on any stages of this years team cup - it was released after team cup already had started

Edited September 21, 2015 by websmile

[viper-rd]

Websmile.

Look at the other results - and explain to me - why they did not HM87/HM86, and not removed?

Only the first three results are correct?

http://ark.intel.com/products/78945/Intel-Pentium-Processor-3558U-2M-Cache-1_70-GHz?q=3558U - skylake??? ))))))))))

[Marquzz]

Why can't I choose processor model in the drop down when entering hardware? It works for other hardware like graphics card and motherboard, bnut not cpu. I used to be able to submit anyway by just writing the exact same as in the dropdown, but now when I try the enter "Pentium 1 90MHz" it just won't except it. And as mentioned, I can't choose that in the drop down either...

 

Using latest Chrome.

[Marquzz]

And I missed uploading result for Via SuperPi 1M since the stages are ending different times! PLEASE DON'T DO THAT AGAIN! The eSport site doesn't even work damnit so it REALLY hard to actually see when stages end!!!!

[websmile]

@viper\-rd As said, I am no expert for this, but since when is skylake 22nm?^^

http://www.cpu-world.com/CPUs/Pentium_Dual-Core/Intel-Pentium%20Mobile%203558U.html

And on the results I checked, all have H86 or H87, you simply google the mainboard/Notebook model and find out which chipset is used. Now please accept that we do our job, the results are OK, which yours was not, if you think your chipset is H86/87 make cpuz/aida whatever screenshot and send to Pieter

Edited September 22, 2015 by websmile

[viper-rd]

Websmile.

Tell me - what my result is different from the second, in addition to the CPU used? Chipset same?

http://hwbot.org/submission/2969111_viper_rd_superpi___32m_pentium_3558u_18min_11sec_14ms/

http://hwbot.org/submission/2949927

 

The screenshot of this result there are no information on the chipset - but this result is not taken away - why?

http://hwbot.org/submission/2934483

[trodas]

if you think your chipset is H86/87 make cpuz/aida whatever screenshot and send to Pieter

 

Not screenshot, whole hardware report will be in order there.

 

While it is not entierly impossible, that some software identify certain chipset(s) wrongly, the error can be corrected, if it is indeed error. But things like that usually happen only on very rare situations on somewhat obscure and less known, old hardware. Take this find for example:

http://forum.hwbot.org/showthread.php?t=143722

 

See, how you can present your case to be dealt with? Done AIDA 64 report, show screens, show the reports and come with other sources, claming the different chipset and then the error get fixed.

 

Just claiming that mainboard X use chipset Y just will not quite cut it.

 

 

 

viper-rd -

The screenshot of this result there are no information on the chipset - but this result is not taken away - why?

http://hwbot.org/submission/2934483

 

Maybe because the Asus G551JM is easy to be found NTB:

http://www.asus.com/cz/Notebooks/G551JM/specifications/

 

With clear specifications:

Intel® Core™ i7 4710HQ procesor

Intel® Core™ i5 4200H procesor

Intel® HM86 Express Chipset

 

...that match the user hardware well? Hardly he could be punished for the fact, that CPU-Z 1.72.1 does not recognize the chipset on this particular NTB.

Edited September 22, 2015 by trodas