Jump to content

OFFICIAL: M17x-R3 Benchmarking Thread


Recommended Posts

hahaha man i don't know what i'm talking about anymore, i need to stop wake and baking. hahahaha

lol its cool man you've got me to vape my brain away too :P

Your 107baseclock overclock was still really sick if you had 3 cpuz you could enter hwbot for highest baseclock on m17x R3 ;) lol

wake and bake... She'll get ya ;)

Link to comment
Share on other sites

yes i know about it, i was one of the very first to see it and test it. i just didn't see enough gains to use it everyday. maybe if i'm encoding something but other then that i don't think it worth it. i stop ocing my M18x cuz it wouldn't boot with oc on battery :(.

man check out this wprime of my old 2720, and this is before xtu, just pure bclk oc.

[ATTACH=CONFIG]3640[/ATTACH]

how did u get this thing to boot like this? i can set the bclk to max 105.5 and even 105.6 intel xtu will freeze the whole system on apply. -.-

Link to comment
Share on other sites

how did u get this thing to boot like this? i can set the bclk to max 105.5 and even 105.6 intel xtu will freeze the whole system on apply. -.-

its was just a beastly 2720, its was actually stable to around 107.6, each cpu is gonna be different my newer cpu only stable to around 104. :(

Link to comment
Share on other sites

how did u get this thing to boot like this? i can set the bclk to max 105.5 and even 105.6 intel xtu will freeze the whole system on apply. -.-

Sorry deadbydawn its luck of the draw. If your benching you can disable extra devices via the bios like bluetooth etc and see if you can boot a higher baseclock but most likely you will see no gains. :(

Like i said its just a matter of the motherboard. My last motherboard was better than last 2 and current one... Only one i had that went to 106.2 baseclock :) this one starts giving me trouble at 104.2-3 if yours maxs around 105.3 and stuff its not so bad you have a nice range to overclock with... If its not working there just go down by a step or two till it works. :) sorry man ILB was only board ive seen at 107mhz i miss my last motherboard too :) but it had to be replaced it had a broken memory slot.

  • Thumbs Up 1
Link to comment
Share on other sites

its was just a beastly 2720, its was actually stable to around 107.6, each cpu is gonna be different my newer cpu only stable to around 104. :(

Yeah exactly... but honestly a little less about cpu than motherboard because of the differences ive seen from motherboard to motherboard... I think certain mobo have more clocks locked like icc and pcie... common limits are 104mhz range 105mhz and rare is 106mhz... With ILB having a golden mobo sample... And probably a decent cpu sample too... if you had xm i have feeling you still would be limited to 105mhz range deadbydawn... Even other alienware laptop owners see same thing ... Ive seen people have trouble at 102-103mhz so we are still lucky in that sense.

  • Thumbs Up 1
Link to comment
Share on other sites

how did u get this thing to boot like this? i can set the bclk to max 105.5 and even 105.6 intel xtu will freeze the whole system on apply. -.-

if you tried higher flex voltage and tried to push higher do you have same results? If not you may find gains and that says its cpu...but its likely not going to help. Id say stay cooler so cpu requires less volts... again though it may not help :(

Have you tried anything else in the bios deadbydawn? Getting used to the way the unlocked menus work?

Link to comment
Share on other sites

if you tried higher flex voltage and tried to push higher do you have same results? If not you may find gains and that says its cpu...but its likely not going to help. Id say stay cooler so cpu requires less volts... again though it may not help :(

Have you tried anything else in the bios deadbydawn? Getting used to the way the unlocked menus work?

thanks so much for your help always! it's a lot of new information, some of it i have to understand first :) i don't know what flex voltage is =/

i disabled watchdog and what i want to try is if i can get past the 105.5 bclk using just the bios. but what i wanted to ask first is, if i set it too high, will the bios then reset itself in order to boot again, or is that only using watchdog? because i don't want to risk a non booting system because of a too high of a bclk...

yeah i got used to it. i just don't see why there are some settings which are found twice in the unlocked bios. but i like it a lot, i always like when something gives me the ability to do something which a system is able to do, but was restricted by the manufacturer in the beginning. :)

Link to comment
Share on other sites

The second menus are important on xm more because multipliers and stuff reset if we use over 45x so we have to change voltage and stuff via another menu or we have to re set the multis again via xtu. Your welcome. O call it flex voltage because its both... But you will see it in the bios called "turbo flex vid" vid as in voltage you see like in throttle stop or cpuz common is 1.22v for that kind of cpu im sure it registers higher... im not sure though if non xm will see higher voltage though... so keep temps low so voltage required stays down.

Disabling that watchdog is actually how it is by default and no if baseclock is too high to boot either it will reboot and have reset to 99.8 or it will not boot but simply resetting cmos battery will reset baseclock to 99.8 everytime as a failsafe. yeah i usually had better luck pushing baseclock via bios over xtu but the motherboard i had that did 106.2 did it by xtu in windows and by force restart.... But all others booted higher via bios.... So i had a 106.2, 105.1, and currently a 104.4 about :| anyway you get the picture find your max but dont fret over it as if your going to use it all the time id drop a few from your max baseclock Intel says 105 and damage or issue as early as 102mhz and more common at 104 (pci/usb/sata clocks) can occur. Seems you have a above average motherboard :)

Link to comment
Share on other sites

so i can't get higher than 105.5 bclk :) but it's okay for me. did a wprime 1024 with bclk 105.0. result is fine with me :)

i won't be doing much more oc'ing atm. i think with over 17k in vantage i pretty much reached my peak ;) will concentrate mor on gaming i believe hehe.... vita needs some attention also ;)

damn, it got warm here suddenly! =/ hate how the weather changeds from one day to the next like 180degrees...post-1903-14494992742019_thumb.jpg

Link to comment
Share on other sites

Here is my best run, not sure im gonna get anything better at this point till i snag a 2920 eventually. :P

<a href="

http://s989.photobucket.com/albums/af16/bigtonyman/Benching%20Stats/?action=view&current=Capture-3.png" target="_blank"><img src="http://i989.photobucket.com/albums/af16/bigtonyman/Benching%20Stats/Capture-3.png" width=600 length=480 border="0" alt="Photobucket"></a>

Edited by bigtonyman
Link to comment
Share on other sites

thx dude :)

i noticed something strange: how can i get my hyperx 1600 to run at the speed they're made for? if i set the ram settings in the advanced performance menu to 1600 and leave everything else, it shows them running at ~699MHz in CPUZ. if i set to one of the xmp profiles 1 or 2 (don't remember the exact name), it will crash sooner or later (i guess because the timings built into those profiles are set way to low). but if windows boots successfully i can see that they run at ~800MHz (for as long until windows crashes BSOD'ing). but if i set it to custom profile and set all the timings manually, it will run fine BUT will show ~699MHz..

Link to comment
Share on other sites

For me it works like this they need over 1.5v they arent low volt ram modules. Set 1600mhz manually, set 1.6 or 1.65v manually and also set it to xmp2 restart with save via f10 and re-enter bios confrim all three stayed... If the volts reset set them back to 1.6v or 1.65v and once this works if you ever use xtu to adjust anything i suggest taking the extra time to confirm that all three ram settings stayed upon reboot. Xm setting multis can reset my ram setting... non xm may still have similar probs. Hope that helps deadbydawn... i have mine at that now 1600mhz manually set 1.6volts ram and XMP2 (xmp1 is for 667x2 at low timings and XMp2 is for 800mhz x2 at low timings) if no xmp is set 1600mhz doesnt load for me and 667x2 default timing instead defaults to 9-9-9-24. So xmp1 is like 7-7-7-21 and xmp2 is like 9-9-9-27

1333mhz stock timing may only take 1.5v any xmp will probably need higher volts... And if. Your still using baseclock overclocked you definately will need more than stock volts even more so on xmp2 1600mhz... You apply 105 baseclock and the xmp2 its probably running about 1640+mhz and will for sure need 1.6v or higher

Edited by mw86
  • Thumbs Up 1
Link to comment
Share on other sites

thx a lot for your infos! i tried as you stated. it worked, but only until i did some gaming. bond crashed the whole time while h.l. lost coast gave me a BSOD linking to a .sys file. i set the ram settings back to as it was before and now it's fine, no crashing during games. but its @~699MHz :( daium.

btw: and i didn't oc the bclk btw!

Link to comment
Share on other sites

thx a lot for your infos! i tried as you stated. it worked, but only until i did some gaming. bond crashed the whole time while h.l. lost coast gave me a BSOD linking to a .sys file. i set the ram settings back to as it was before and now it's fine, no crashing during games. but its @~699MHz :( daium.

btw: and i didn't oc the bclk btw!

Dont give up my friend they are guaranteed to 1600mhz and fully function... we will get to bottom of this worse case you determine you have a single bad ram stick or a pair and have to have them send you a replacement set... :) they are guaranteed so keep your stuff in case you need to return...

okay wasnt sure about baseclock.... since you were benching earlier... no probs... doesn't make sense only thing that can limit is the cpu as 1333mhz may be max of the cpu but m14x users have used 1866mhz plug and play ram and had it run full speed :( you shouldnt get those bsods? If volts weren't enough there are some common errors

http://forum.techinferno.com/off-topic/1124-general-overclocking-advice-record-breaking-benchmarks.html

*I believe you did all correctly and maybe there could be a faulty ram module as yours are rated to 1600mhz. since your having issues set 1.65v,1600mhz, and XMP2 and ensure all show correct in XTU and CPUZ (dont read ram volts from cpuz inaccurate) before testing in any games or apps. If all show correct test in newer games please that youve never had issues in I can say any half life is older and you might just be having a error i had some a few months ago loading older games back from directx9 and lower era... what modern title do you have... you play bf3, skyrim, or anything else... idea is use a modern game that also pushes harder than Half Life. I could suggest using memtestx86 on a bootable cd.

Deadbydawn the errors happened on which settings specific?

*i hate to even mention but re-seating memory can sometimes resolve tons of issues but i hate that suggestion because its usually beat to death when solving issues...

*did you confirm voltage showed correct in XTU? on the right hand section if it says ram volts standard and current the current value is accurate... i get bsods if my volts were reset to 1.5v for 1600mhz XMP2.

What type of settings are you using that arent default in the unlocked or stock bios? (which performance or non performance changes have you been using)

can you show me the timings and ram speed in cpuz in 1600mhz xmp2? you can just open cpuz twice and open the two tabs.

....either way let me just get an idea of what things are set like for you because you absolutely don't need to accept that your ram isnt working at 800mhz dual channel (1600mhz) rated for that at 9-9-9-27 timings at at most 1.65v. If we cant solve this you should determine which sticks are causing issue and seek a replacement from Kingston. If you get to this point be sure to try each stick individually or in their matched pairs in the dimms 0 and 1 of the notebook. Utilizing memtestx86 or testing with cpu tests ive had plain cpu tests reveal ram failures like wprime. easiest is memtestx86 :) i despise it though as for most it can take any wheres from immediately to a hour or two to reveal ram errors :(

hang in there Deadbydawn we will tell that ram who is boss or find the fault of the errors :)

good luck let us know if you need any help :)

these aren't set in stone some memory errors and cpu errors can cause other issues which will show as different errors in the effected program.

BSOD Codes for SandyBridge

0x124 = add/remove vcore or QPI/VTT voltage (usually Vcore, once it was QPI/VTT)

0x101 = add more vcore

0x50 = RAM timings/Frequency add DDR3 voltage or add QPI/VTT

0x1E = add more vcore

0x3B = add more vcore

0xD1 = add QPI/VTT voltage

“0x9C = QPI/VTT most likely, but increasing vcore has helped in some instances”

0X109 = add DDR3 voltage

0x0A = add QPI/VTT voltage

Edited by mw86
  • Thumbs Up 1
Link to comment
Share on other sites

Dont give up my friend they are guaranteed to 1600mhz and fully function... we will get to bottom of this worse case you determine you have a single bad ram stick or a pair and have to have them send you a replacement set... :) they are guaranteed so keep your stuff in case you need to return...

okay wasnt sure about baseclock.... since you were benching earlier... no probs... doesn't make sense only thing that can limit is the cpu as 1333mhz may be max of the cpu but m14x users have used 1866mhz plug and play ram and had it run full speed :( you shouldnt get those bsods? If volts weren't enough there are some common errors

http://forum.techinferno.com/off-topic/1124-general-overclocking-advice-record-breaking-benchmarks.html

*I believe you did all correctly and maybe there could be a faulty ram module as yours are rated to 1600mhz. since your having issues set 1.65v,1600mhz, and XMP2 and ensure all show correct in XTU and CPUZ (dont read ram volts from cpuz inaccurate) before testing in any games or apps. If all show correct test in newer games please that youve never had issues in I can say any half life is older and you might just be having a error i had some a few months ago loading older games back from directx9 and lower era... what modern title do you have... you play bf3, skyrim, or anything else... idea is use a modern game that also pushes harder than Half Life. I could suggest using memtestx86 on a bootable cd.

Deadbydawn the errors happened on which settings specific?

*i hate to even mention but re-seating memory can sometimes resolve tons of issues but i hate that suggestion because its usually beat to death when solving issues...

*did you confirm voltage showed correct in XTU? on the right hand section if it says ram volts standard and current the current value is accurate... i get bsods if my volts were reset to 1.5v for 1600mhz XMP2.

What type of settings are you using that arent default in the unlocked or stock bios? (which performance or non performance changes have you been using)

can you show me the timings and ram speed in cpuz in 1600mhz xmp2? you can just open cpuz twice and open the two tabs.

....either way let me just get an idea of what things are set like for you because you absolutely don't need to accept that your ram isnt working at 800mhz dual channel (1600mhz) rated for that at 9-9-9-27 timings at at most 1.65v. If we cant solve this you should determine which sticks are causing issue and seek a replacement from Kingston. If you get to this point be sure to try each stick individually or in their matched pairs in the dimms 0 and 1 of the notebook. Utilizing memtestx86 or testing with cpu tests ive had plain cpu tests reveal ram failures like wprime. easiest is memtestx86 :) i despise it though as for most it can take any wheres from immediately to a hour or two to reveal ram errors :(

hang in there Deadbydawn we will tell that ram who is boss or find the fault of the errors :)

good luck let us know if you need any help :)

these aren't set in stone some memory errors and cpu errors can cause other issues which will show as different errors in the effected program.

BSOD Codes for SandyBridge

0x124 = add/remove vcore or QPI/VTT voltage (usually Vcore, once it was QPI/VTT)

0x101 = add more vcore

0x50 = RAM timings/Frequency add DDR3 voltage or add QPI/VTT

0x1E = add more vcore

0x3B = add more vcore

0xD1 = add QPI/VTT voltage

“0x9C = QPI/VTT most likely, but increasing vcore has helped in some instances”

0X109 = add DDR3 voltage

0x0A = add QPI/VTT voltage

naah, i won't give up :) thanks once more for your premium help! :)

so, i reset the bios loading the optimal defaults. disabled bluetooth and card reader (never use this ish anyways). then i set the memory override support to enabled, memory voltage to 1.65 volts and dimm provile to xmp profile 2.

then i got already a question: xtu always tells me that it couldn't apply some of the settings and sets the turbo boost power max back to 45w. should i set it in bios like this (as it is a 45w cpu anyways):

long duration pwr limit: 45

l.d. time window: 56

s.d.pwr limit: enabled

s.d.pwr limit: 72

because it anoys me that each time i start xtu it's telling me that.. oh and if i set anything in xtu, the memory override support is disabled after restart and i have to enable it again in bios.

i made two screens, hope they help:

post-1903-14494992759077_thumb.jpg

post-1903-14494992759852_thumb.jpg

what i just noticed is, that in cpu-z it says max bandwidth "pc3-10700 (667MHz)" but the part number shows it right, as they are 1600MHz ones.

i just ran a memory stress test in xtu which failed after around 12 minutes. i will take out two of the ram modules now and then run the test again. afterwards doing the same with the two other ones.

  • Thumbs Up 1
Link to comment
Share on other sites

second pair also passed. switched the ram to the harder-to-reach slots now. doing the same thing again. hmmm

btw, i'm thinking of removing 8GB of ram anyways, i never ever need 16GB. but this is just a side-note ;)

the second pair in the harder-to-reach dimm slots also passed 20min. i will test the first pair now. then again both pairs at once. i know 20 min is short for memory testing, but i thougt since before, the test failed after like 12 minutes already maybe i can determine the error quickly.

i noticed that some times when i start cpu-z, the register SPD is empty, meaning in the register memory it will show there are (atm) 16GB of 1600MHz memory installed. then in the register SPD i can browse through all of the 4 dimm slots, but everything is shown empty. then after a restart everything is shown correctly again.

now doing all 4 dimms used with the whole 16GB again (test).

Edited by deadbydawn
Link to comment
Share on other sites

so i ran 30min. with 16GB installed. passed. now i ran 20min. it failed due to thermal throttling. when i looked at the monitoring in xtu one of the cpu cores hit 82degrees and then a throttling of 1% happened, so the test aborted. so this time, thermal throttling aborted the test. i didn't use throttle stop d'oh. i will do so right now and try again :) strange thing though, as the cores hit 82 before and i didn't have throttling.

this time it failed. throttle stop was running and no throtteling. damn i start to be puzzled =/

what i discovered is, that i never set the ram settings in xtu. i now used the 1600mhz preset and adjusted the voltage to 1.65 as in the bios. rebooted. bios settings (memory override support) weren't overwritten. i did a 30 minutes test and now a 20 minutes one is just about to finish. will do another one right after this one is done.

Edited by deadbydawn
Link to comment
Share on other sites

so i ran 30min. with 16GB installed. passed. now i ran 20min. it failed due to thermal throttling. when i looked at the monitoring in xtu one of the cpu cores hit 82degrees and then a throttling of 1% happened, so the test aborted. so this time, thermal throttling aborted the test. i didn't use throttle stop d'oh. i will do so right now and try again :) strange thing though, as the cores hit 82 before and i didn't have throttling.

this time it failed. throttle stop was running and no throtteling. damn i start to be puzzled =/

Your doing great cpuz and bios check out set them like you did. For now stop using XTU altogether till we solve this. Make a memtestx86 bootable media i see no other way to confirm your ram accurately. The XTU memtest seems to do too much with cpu also and is messing up your testing as you noticed because it forced you to use throttle stop... So now do the memtestX86 as it involves no windows at all. Get the bios set like in your pic do full restart and confirm it says exactly what you showed in bios (btw your doing great) (ie 1.65v,1600mhz,xmp2)

boot from the bootable media and test 8gb at a time with memtestX86 otherwise we dont know the which set is giving problems.... Look on the side of mem stick there should be a unique number on each stick keep matched pairs together its also pretty important. Example a set has 801 and the pair for it has 802 and my other pair of 8gb might have 321 and 322... so try to see that when testing the pairs its the one that came with the other one...

Now only test 2 sticks at a time in a pair so its dual channel 8gb 1600mhz...

*your ram is showing in cpuz like mine no faults at all. Mine says 1.5v via cpuz to and says the chips are 667mhz also... They are 1333mhz modules that are rated to 1600mhz at xmp2 timings at 1.65v :) so no issues but yes they are by default simply 1333mhz chips that are rated upto 1600mhz. The dimms you put the ram in for just two sticks are the right ones disregard where i said dimm 0 and 1... (just ensure they are next to each other in pair not 1 in one set of slots and another in the other pair of dimms... of which i already know you arent doing that so continue on as you have)

Strange yes that you reveal failure when they are set correct yet other times they gave no issues... So round three hopefully something is revealed or they just start working. If you decide to use only 8gb at 1600mhz thats fine but pick the fully working repair and return the non functioning one or get replacement pair from Kingston and sell them to someone :) so again drop use of XTU till you solve this and only work on this via bios settings... :) ill keep looking for your progress. Btw cant believe how much you did try over night :) lol you work fast ;) catch you later good luck i hope something is determined this next round :) and i hope its no problem running memtest for our next test. Seems if you run each pair of 8gb on the extended testing pattern one set if faulty is bound to show an error in a few hours max... Set the test going and tinker with other stuff if nothing shows error in first few minutes... Set it and forget it.

  • Thumbs Up 1
Link to comment
Share on other sites

hehe, well, what's at night for you is day for me ;-) i'm working in the IT so i do some work and besides do my own stuff i.e. fix the m17x haha before i would just stay up all night doing this, but now with an upcoming wife (saying i'm getting married) i don't think she would be too pleased about me spending all night the whole night just geeking off hahaha ^^

thx so much for your advice! the last 5 memory stress tests i did with xtu were all passed. so now i created the bootable memtest86 cd and it's firing. it just started right away when it booted. i will let it run for an hour, then pack up my stuff, go home and then do the tests as you suggested, 8GB per time. i will probably just let it run over night afterwards then.

strange thing about my hyperx is, the unique number on the 4 is the same. i looked at them when i was taking them out, because i was trying to do exactly what you stated a moment ago: match the pairs. but i couldn't figure out at all which belongs to which. same unique number. and then there is a long-@ss number besides which is completely different on all four (not even a hint on same structure or whatsoever).

we'll see what round 3 will show :D

Link to comment
Share on other sites

so it showed an error. i now took the two hyperx out of the dimm 3&4 and am testing 8GB in slot 1&2.

once more i looked at the memory but those unique numbers didn't give me a hint. i was mistaken, the numbers are not the same, they are completely different, all 4 of them. then again there are other numbers (probably part numbers) which are exactly the same. well, i'm doing 8GB-tests now and then the other 8GB. hope it finds an error again. would really blow if in the end it was a socket which was f*cked or something like that. rather have an error, put in the other 8GB and thos be fine. :) then i could just send the broken ones to kingston for a replacement.

so the first set of 8GB passed the first run on all 11 tests without errors. i put in the second set now and am running the tests. if nothing is found i will at home let each of the sets run for a couple hours to see what happens then. hmmmmmmmm

Edited by deadbydawn
Link to comment
Share on other sites

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...

Important Information

By using this site, you agree to our Terms of Use. We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.