Jump to content
HWBOT Community Forums

Recommended Posts

  • Replies 76
  • Created
  • Last Reply

Top Posters In This Topic

Posted
How am I supposed to get a Data File for Hwbot Prime on Mobile...?

 

For HWBOT Prime on mobile, you'll need to submit via the app.

 

I can not put the result stage 2 on ASUS P5N32-E SLI Plus.

This board has North Bridge NVIDIA nForce 650i SLI SPP rev A2.

but here it is http://hwbot.org/hardware/motherboard/p5n32_e_sli_plus/ --- NFORCE 4 SLI X16 ???

Please fix this error.

here is an example CPU-Z: http://valid.canardpc.com/lejs7d

thank

 

Fixed :)

  • 2 weeks later...
Posted

Question about the longest 1B run.

 

Do you have to have uniform speed on the run for it to be alowed or would it be ok to remove some load after lets say 15 batches to speed it up in the end if it takes to long?

 

how exact does the run need to be, would a run with an old motherboard without HPET that got a slight rounding error be valid? Aka a run where the batch times looks like the following example, yes I know its not valid or even complete like the screenshoot is now but wonder on the principle of the times.

0756f1bbc1.png

Posted

I'm fairly sure that if you get that bloody big red 'X' that it won't be accepted. If you don't, it's all good for submission no matter how you got there, speed up, slow down, its all good.

Posted
No, that's not necessary. GPUPI 2.2 indeed fixes a bug for these slowest result competitions, but it's nothing that can't be moderated manually. Although we will have a good look at slowest results made with GPUPI 2.1.x in the currently running team cup stage. That's for sure. ;)

 

If there is a bug that affects this competition shouldn't it be announced so that no one is inadvertently (or not) using it to slow the calculation. Especially since it can be moderated easily enough, or is this the type of bug that is more of a hack. I'm assuming that running stuff in the background to slow the calculation isn't considered a bug.

Posted

Slowing the already slow CPU by some settings are not forbidden or against the rules. The validity of resulting file counts most as well, as regularity of the Batch times, IMHO. Yes, I spoted one exploitable bug and I reported it ASAP and it cannot be publicaly mentioned right now. But I have no need to use it to slow down my score anymore :)

 

It is slow-enought, or is not it?

 

GPUPI_324h_and_counting_on.jpg

 

:D :D :D

Posted

There's public API that allows to pause executing code. Also using a debugger (ida,wdasm etc.) you can simply step-in/pause, controlling each loop easily. Oh well, can't wait to get back from holiday and start the machine up.

Posted

Well, I prefer hardware way to do the slow-downs. In some cases it works great (Aquamark), but on other cases it produce mixed results (SuperPi 32M). Disabling L1 cache/L1+L2 caches on fully stable mainboards cause the SuperPi always fail on 32M test during it, tested on 3 different machines (Jetway V266B, 100x5 AXP-M, 2048MB 2-3-2-6 DDR1 Qimonda rams - PCchips M810LR, 100x11 Barton 2500+, 256MB 3-3-3-6 SDRAMs - MSI PM8M3-V, 200x17 Pentium 4 650, 2048MB 2-3-2-5 DDR1 OCZ rams):

 

Super_Pi_crashed_when_run_without_caches.jpg

 

...but that is a bit OT there (that is aimed at 12+h SuperPi 32M runs)... yet if you have some infos, pls drop me a PM or start a thread about this...

Posted
There's public API that allows to pause executing code. Also using a debugger (ida,wdasm etc.) you can simply step-in/pause, controlling each loop easily. Oh well, can't wait to get back from holiday and start the machine up.

 

Okay, I see. Those actually interact with the actual program/code though. I can see why that would be banned.

 

Does it mean that non-even loops will cause a sub to be disqualified? I know the current run I have has them uneven because I turned off the background program(s) to make sure it was actually running or had frozen up. I wish it wasn't so vague on what the rules are for this type of stage. A ten or twenty day run is a lot of time to have to rerun because of unclear rules.

Posted

Why are we splitting HWBot Prime for Pi up into a per core score? There was no mention of this previously, so in essence you will need a single core to win this which is just stupid since prime doesn't scale well with cores.....

  • 2 weeks later...

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...