You are not registered yet. Please click here to register!
The Rampage II Extreme Thread - Page 2 - i4memory.com - different look at memory

Notices

Reply
 
Thread Tools
Old 06-02-2010, 12:00 AM   #9
CryptiK
Senior Member
CryptiK's PC Specs
 
Join Date: Aug 07 2008
Posts: 774
If you look a few posts up, I ran Prime small FFT's for 3 hours at 218 BCLK, and I think it would have probably passed Linx too, in my testing the only thing LinX requires over prime is a little more vcore to counteract the increased CPU temps.

I was using PCI-E 100MHz for 218 BCLK, never really tested higher as I didn't need to.
CryptiK is offline   Reply With Quote
Old 06-02-2010, 12:09 AM   #10
Grinchy
Junior Member
Grinchy's PC Specs
 
Join Date: Feb 05 2010
Posts: 10
Hi .
LinX 8 threads is nowhere close to IBT xtreme mode 32 threads.
To give you an idea i passed 19 pass of ibt(that uses linpack just like linx) 8 threads with about 5 GB of memory but then failed at the first pass(system freeze) when using 32 threads.
The amount of voltage bump needed depends upon the situations.
For example i noticed that bumps needed are more consistent when using high memory speeds.
I used Prime too in the beginning,but IBT is more stressing and more reliable,and quicker too
Anyway what do you think about the wall i have.Isn't it a too poor result for this mobo?

Last edited by Grinchy; 06-02-2010 at 12:18 AM.
Grinchy is offline   Reply With Quote
Old 06-02-2010, 12:26 AM   #11
Grinchy
Junior Member
Grinchy's PC Specs
 
Join Date: Feb 05 2010
Posts: 10
Anyway yours seems an excellent result,in the hardware canucks review that i read they got upto 210 bclk.
Given that your mobo seems willing to get high,could you try the test in the way i told?i am really curious
Grinchy is offline   Reply With Quote
Old 06-02-2010, 12:28 AM   #12
CryptiK
Senior Member
CryptiK's PC Specs
 
Join Date: Aug 07 2008
Posts: 774
Interesting I've never tried IBT with 32 threads. I used the original one ages ago then switched to using LinX.

I do think 200 BLCK is a little low for the board, perhaps it could be your CPU? Have you tried using a lower multi and just testing BLCK?

EDIT - for sure, I'll DL it and have a go. Can't run it tonight as its already pretty late, but I'll do it over the weekend and see how it goes. Do I just use the latest version of IBT and set it to 32 threads (to replicate your settings)?

Last edited by CryptiK; 06-02-2010 at 12:40 AM.
CryptiK is offline   Reply With Quote
Old 06-02-2010, 12:47 AM   #13
Grinchy
Junior Member
Grinchy's PC Specs
 
Join Date: Feb 05 2010
Posts: 10
Remember to right click on the start button in the main window and select xtreme stress mode.I advice you to log,and the reason why i tell 20 pass is that once happened that it crashed after 14-15,so i think that 20 is a good margin.

Edit f course i tried lower multipliers,trying to avoid multiple of two.
FOr example .as i told you i operate at 200.I am 4.2 Ghz 1.375 Vcore,1.375 cpuvtt,1603 ddr3 @1.65volt,stable like rock.
But if i move to 201 as i start ibt it freezes ,no matter how much i bump voltages.
Lowering multiplier things don't change,that's why i blame the mobo.

edit: that's why i say yours is an excellent result

Last edited by Grinchy; 06-02-2010 at 12:59 AM.
Grinchy is offline   Reply With Quote
Old 06-02-2010, 05:28 PM   #14
CryptiK
Senior Member
CryptiK's PC Specs
 
Join Date: Aug 07 2008
Posts: 774
I had a go with IBT, now I remember why I stopped using that program! It's so inefficient, you load the system with 8 threads and max ram and the system becomes unresponsive during testing then finishes after some time. I don't have time to muck around with it until next week but I think if you are prime and linx stable you'll be able to pass anything else you need.
CryptiK is offline   Reply With Quote
Old 06-02-2010, 09:40 PM   #15
Grinchy
Junior Member
Grinchy's PC Specs
 
Join Date: Feb 05 2010
Posts: 10
errr...this is why it is the most reliable system stability test, because it loads the sytem completely.The CPU is completely loaded by the test.Nothing like it loads completely the CPU,that's why it is so unresponsive.It uses linpack in the same way LinX does but with extreme stress mode and using a high number of threads it makes a more heavy use of it. Using 32 threads is fundamental if you want to check the true stability of the system. As i told you using 8 threads you can easily pass 20 passes and think that your system is rock stable,but the truth is that you have a system that will probably make a lot of errors,during heavy applications.I was stable for 19 pass with 8 threads but it freezed at 1 pass with 32 threads.I was using xtreme stress mode(that loads the system even further).If you wanted to simulate a LinX run you shouldn't use Xtreme stress mode and use only 8 threads,guess how much LinX can be reliable....(expecially with only ten passes). The fact that it is unresponsive doesn't mean that it is not efficient.How are you going to completely saturate and test your cpu if the test leaves you enough resources to do other tasks?
Prime stability test takes much time for this reason:it doesn't load completely cpu,it uses a limited number of registers.so you must hope that after a long time(not less than 10 hours) all registers have been tested(it is random whether they will be all tested ,but after this long time it is likely).By saturating them all,with IBT(that anyway is the same test that intel use for their cpus...) after a much shorter time(not more than a couple hours with the maximum settings and a 4 Ghz Cpu) u have the warranty of a stability that with prime you could never have.With prime u can say that after a long testing it is probable (but not sure) that all of your cpu will have been tested.WIth IBT you are sure!
With prime you cannot reallysay anything before 10-12 hours.and we are far from warranties that you have after two hours of IBT.Personally i had crashes after 4, 5, 8, 9 hours of Prime95.Never had after 10 hours,but again u are trusting the fact that after such long time it is PROBABLE that your cpu has been completely tested,not relying on somehting sure.LinX,using the same amount of time of Prime95,is a bit better but as i realized it is not enough customizable (not more than 8 threads and no xtreme stress mode).

PS : i didn't want to go off-topic,just wanted to propose a stability system methodology that doesn't give false assurances,but that makes sure that we have a system as stable as a stock one.Would you use Matlab with an overclocked system that passed only 10 pass of 8-threaded linX? i advice you not to...if you don't want to discover that 2+2=3

PPS : don't use real temp with IBT. It will slow down IBT in a disgusting way.I think it is because RealTemp takes system resources, that ibt keeps on waiting to obtain, to measure temperatures.In this way ibt can always remaining waiting and it will never really start testing.I Used lcd poster to check temps

Last edited by Grinchy; 06-02-2010 at 10:18 PM.
Grinchy is offline   Reply With Quote
Old 07-02-2010, 12:20 AM   #16
CryptiK
Senior Member
CryptiK's PC Specs
 
Join Date: Aug 07 2008
Posts: 774
I haven't tested the association with realtemp, but I was using it at the same time so it may partially explain why it was so unresponsive. I am all for testing stability, but within reason too I suppose, and I don't run anything so intense that absolute accuracy is required. However IIRC CPU's have certain errata which will cause errors under certain conditions even at stock speed and voltage, so to my knowledge we can never have a totally accurate processor. It would be nice to get it stable with IBT max stress mode though.

I'll give it a go with 32 threads etc, but when I have the time during the week.
CryptiK is offline   Reply With Quote
Reply

Bookmarks

Tags
extreme, rampage, thread


Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools


Similar Threads for: The Rampage II Extreme Thread
Thread Thread Starter Forum Replies Last Post
Asus Rampage III Extreme - results thread eva2000 Asus Intel motherboards / CPU 484 07-11-2011 02:57 AM
RAMPAGE III EXTREME REVIEW Uncle Fester Asus Intel motherboards / CPU 5 06-05-2010 03:44 AM
[Techpowerup.com] (PR) ASUS Launches the Rampage III Extreme and Maximus III Extreme Motherboards News Reviews & News Online 0 16-04-2010 11:40 AM
i5 670 & ASUS P7P55D-DELUXE Motherboard 2D review Uncle Fester Asus Intel motherboards / CPU 9 18-03-2010 07:41 AM
Asus Rampage Extreme: having fun with MCH, Spi under Air and Cascade coolings before Asus Intel motherboards / CPU 2 08-12-2008 05:24 AM


All times are GMT +11. The time now is 08:14 AM.