The Corsair User Forums  

Go Back   The Corsair User Forums > Corsair Product Discussion > Memory

Reply
 
Thread Tools Search this Thread Rating: Thread Rating: 53 votes, 4.96 average. Display Modes
  #16  
Old 12-05-2008, 10:37 PM
syscev syscev is offline
Registered User
 
Join Date: Nov 2008
Posts: 30
POST ID # = 379298
syscev Reputation: 10
Default Reply 2..

Quote:
Originally Posted by Corsair Technical Support
I would not suggest the Voltage over 1.7 Volts but you are welcome to try what ever ou like.
This is hilarious stuff ! Ooohh.. The contradictions...
Reply With Quote


  #17  
Old 12-06-2008, 01:24 PM
DerekT's Avatar
DerekT DerekT is offline
Registered User
DerekT's PC Specs
 
Join Date: Jul 2005
Location: White Rock, British Columbia
Posts: 10,168
POST ID # = 379338
DerekT Reputation: 10
Default

Quote:
Originally Posted by syscev View Post
There some rule saying the memory voltage should not be more then +0.5V of the Cpu Voltage ?
Yes, if you raise CPI voltage, then it seems that you will be safe if you then concurrently raise Vdimm. This is what is being seen at Xtremesystems. I personally would NOT raise the Vdimm without a concurrent raise of the QPI voltage.
Reply With Quote


  #18  
Old 12-15-2008, 05:52 AM
syscev syscev is offline
Registered User
 
Join Date: Nov 2008
Posts: 30
POST ID # = 380455
syscev Reputation: 10
Default Hmmm... New BIOS for Eclipse

Hmmm... Looks like there is a new bios, 1.3b3... Positive feedback on memory compatibility and OC. I better try it tonight.

http://forum.msi.com.tw/index.php?to...2215.msg922754
Reply With Quote


  #19  
Old 12-15-2008, 01:22 PM
RAM GUY's Avatar
RAM GUY RAM GUY is offline
Corsair Product Guru
 
Join Date: Apr 2001
Location: Fremont, CA
Posts: 118,269
POST ID # = 380485
RAM GUY Reputation: 10
Default

And please let us know how you make out!
__________________
Support accounts and tickets can be created at https://support.corsair.com.
Reply With Quote


  #20  
Old 12-18-2008, 08:01 PM
will160628 will160628 is offline
Registered User
will160628's PC Specs
 
Join Date: Dec 2008
Posts: 10
POST ID # = 381121
will160628 Reputation: 10
Sad First time Corsair user

I'm hoping he had better luck than I did with the bios update. I've got almost the same issue only worse and with the 6GB kit TR3X6G1600C9. XMP causes my system to lock at CPU init unless I set my bclock up to 166 (weird but true). Now this makes me think that there is a bios issue on this board with Corsair memory. Every memory issue I've seen in my search for a fix has turned out to be on a Corsair+Eclipse combo. Now on top of that issue I may actually have some problems with the sticks I have anyway. Mine are only prime 95 stable at 1333 with timings of 9-10-10-25 1.55v and even then I sometimes get a restart after about 12 hours. Now at 1200 1.50v and auto timings they're rock solid (know its not the processor because i have to overclock it to get to this speed). I wish I could try another brand and send this stuff back but the well known online store i bought it from doesn't do refunds on memory. Oh btw the 1600mhz problem at least on mine hasn't been solve in bios 1.0,1.2,1.3b2 or 1.3b3. I tried them all and did get better stability processor-wise but no change in the memory. Most people do see it resolve with 1.3b2 and beyond though. I'm just not that lucky.

Forgot to mention ran memtest86+ through 6 passes with 1600mhz memory and 160 bclock (cpu I believe 3.2ghz) with no errors and this is on a fresh install of Vista 64 (had the restarting problem on my original install and thought it was OS related).

Last edited by will160628; 12-18-2008 at 08:12 PM. Reason: Added last line
Reply With Quote


  #21  
Old 12-22-2008, 02:41 AM
DerekT's Avatar
DerekT DerekT is offline
Registered User
DerekT's PC Specs
 
Join Date: Jul 2005
Location: White Rock, British Columbia
Posts: 10,168
POST ID # = 381398
DerekT Reputation: 10
Default

Quote:
Originally Posted by will160628 View Post
Forgot to mention ran memtest86+ through 6 passes with 1600mhz memory and 160 bclock (cpu I believe 3.2ghz) with no errors and this is on a fresh install of Vista 64 (had the restarting problem on my original install and thought it was OS related).
You may find that increasing the MCH voltage can bring a stability when both Prime and Memtest pass.
Reply With Quote


  #22  
Old 12-22-2008, 02:27 PM
will160628 will160628 is offline
Registered User
will160628's PC Specs
 
Join Date: Dec 2008
Posts: 10
POST ID # = 381466
will160628 Reputation: 10
Default continued degredation

I actually tried pushing all of my voltages on the board up to just below "dangerous" voltages on Friday and it didn't help. Turned everything back down after it didn't help (didn't push memory past 1.65, I was trying to get it to boot at 1600). Funny thing is late Saturday I had firefox open in the background and tried to open something (don't remember what) and had a sudden restart again only with the memory at 1200Mhz this time. Now its getting harder to keep 1200Mhz stable. I have to admit I've never seen memory gradually get less and less stable like this especially running at 1.5v(actually if I run the power saving feature of my motherboard it runs at 1.3v). From watching it, the memory voltage never seems to go over 1.5 in the monitoring software but I swear this stuff acts like its been over-volted. I went on and ordered some more memory from a different manufacturer so i can try it out (1600/1800 so it should be stable at 1600 no matter what). It has a 30 return policy so I don't have to worry about keeping it if it doesn't fix the problem. Having to run this memory at 1066 now has me a bit worried though. I'm just hoping it can hold out until my new modules get in. I'm also hoping that if the new modules work it will prove to support that there is actually a problem. I got a canned response to update my bios when I submitted a ticket for this. Thanks for the advice. I might try messing with just the MCH when I get in from work tonight. Then again i might just break down and start troubleshooting which module is the problem child of the bunch.
Reply With Quote


  #23  
Old 12-29-2008, 05:08 AM
syscev syscev is offline
Registered User
 
Join Date: Nov 2008
Posts: 30
POST ID # = 382175
syscev Reputation: 10
Default Advise///

Quote:
Originally Posted by will160628 View Post
I actually tried pushing all of my voltages on the board up to just below "dangerous" voltages on Friday and it didn't help. Turned everything back down after it didn't help (didn't push memory past 1.65, I was trying to get it to boot at 1600). Funny thing is late Saturday I had firefox open in the background and tried to open something (don't remember what) and had a sudden restart again only with the memory at 1200Mhz this time. Now its getting harder to keep 1200Mhz stable. I have to admit I've never seen memory gradually get less and less stable like this especially running at 1.5v(actually if I run the power saving feature of my motherboard it runs at 1.3v). From watching it, the memory voltage never seems to go over 1.5 in the monitoring software but I swear this stuff acts like its been over-volted. I went on and ordered some more memory from a different manufacturer so i can try it out (1600/1800 so it should be stable at 1600 no matter what). It has a 30 return policy so I don't have to worry about keeping it if it doesn't fix the problem. Having to run this memory at 1066 now has me a bit worried though. I'm just hoping it can hold out until my new modules get in. I'm also hoping that if the new modules work it will prove to support that there is actually a problem. I got a canned response to update my bios when I submitted a ticket for this. Thanks for the advice. I might try messing with just the MCH when I get in from work tonight. Then again i might just break down and start troubleshooting which module is the problem child of the bunch.
Actually I opted not update to bios v1.3b3. A friend of mine suggested that I don't use the beta bios and rather wait for the official release so, I just going to do that since I have not major compelling reason to update it at the moment.

But after about a month of testing it seems there are still lingering problems at 1600MHz on the 1.3b2 bios. It is stable enough to run windows and do normal work but it is not very stable for gaming.. Fallout 3 crashes after some time, crysis warhead crashes after some time, supreme commander crashes after sometime and far cry 2 crashes after a 5 to 10 mins...

Initially I thought it was the game's problems so, decided to finished up fallout 3 first .. (ARGGHH ! I hate the ending..) Tends to crash intermittently after 30 mins to 1hr+ of game time for no apparent reason.

Now I starting to play Far cry 2, I am willing to test this issue again because I recalled Far cry 2 to be stable before... Clean installing the video drivers and all didn't help.. Then I decided why not tone down the ram speed to 1333MHz and like magic Far Cry 2 plays stable for hours without crashing... Hiaz.. Seems like 1600MHz isn't stable enough for gaming.


Dude you should try setting everything manually.
Frequency 667MHz (Ok multiplier set to 5 for 1333, 6 for 1600)
CAS# Latency 9 (Found this)
RAS# to CAS# 9 (Found this)
RAS# Precharge 9 (Found this)
Tras 24 (Found this)
Trc 41 (Can't find it.)
Command Rate 2T (Set to 2N)
DRAM Volts to 1.65v (Greyed out doesn't mean you can't change it.)

Try 1333Mhz first... Stable enough for gaming...

Last edited by syscev; 12-29-2008 at 09:07 AM.
Reply With Quote


  #24  
Old 12-30-2008, 08:28 AM
syscev syscev is offline
Registered User
 
Join Date: Nov 2008
Posts: 30
POST ID # = 382397
syscev Reputation: 10
Exclamation Ohhh Some updates from MSI...

Would like to point out one thing. Now I managed to mitigate the problem somewhat I don't bother checking their support ticket any more. Only received an emailed noticed today.

They claim that their latest v1.35 bios has resolved some XMP problems. Looks like I must try this out later for the sake of other people... Hope I don't brick my board or something.... Oh well the local distributor like bus ride away... (Happens to be distributor for Corsair too btw )

Here's the support log:

Quote:
Originally Posted by Posted: Dec 02, 2008 by Syscev
I seem to have managed to resolve the problem by setting the configuration manually and not using XMP.

You can refer to this post in the Corsair support forums.
http://www.asktheramguy.com/forums/s...ad.php?t=74660
(I notice the thread number might change you can also go to http://www.asktheramguy.com/forums/f...play.php?f=130
And look for my thread subject "TR3X3G1600C9 + triple channel + MSI Eclipse x58 very unstable")

You can also refer to other complaints of Corsair DDR3 instability at 1600Mhz on triple channel using the MSI Eclipse on this review site.
http://www.hexus.net/content/item.php?item=16331&page=4

Your XMP Profile implementation could be bugged.
Quote:
Originally Posted by Posted: Dec 22, 2008 by MSI
Dear,

Sorry for my late reply.

For your concern,we have done a test.Regrettably here we don't have the memories you mentioned.
The test configurations are as follows:
MB:7520-010(BIOS:v1.34)
CPU:intel i7 920
Mem:DDR3-1600/Corsair/CM3X2G1600C9DHX without xmp(triple channel)
DDR3-1333/********/KHX11000D3LLK2/2GX with xmp(dual channel)
VGA:R4850

As memories of DDR3-1600 don't support xmp,we could not test the issue.But we could install the system normally.

For your concern,please go to our ftp BIOS/7520 to download the biosfile A7520IMS.134 to update then check again.
Now you could install the system normally with triple channel?

Best regards,
MSI technical support team
Quote:
Originally Posted by Posted: Dec 24, 2008 by MSI
Dear,

For your concern,A7520IMS.135 has solved some xmp issue.Please update the bios to it.

Could you please tell us the result of your issue?If the issue remains,we will ask some departments to check it.

Sorry for any inconvenience caused by us and thank you for your cooperation.

Best regards,
MSI technical support team
Quote:
Originally Posted by Posted: Dec 30, 2008 by MSI
Dear,

Could you please kindly tell us whether the issue has been resolved by flashing the bios.

Thank you for your cooperation.

Best regards,
MSI technical support team
Reply With Quote


  #25  
Old 12-30-2008, 10:33 AM
syscev syscev is offline
Registered User
 
Join Date: Nov 2008
Posts: 30
POST ID # = 382408
syscev Reputation: 10
Default Hiaz... See below

It failed... bios v1.35 doesn't help the xmp issue.... Doing intel burn test v1.9 at 1600mhz (manual) is also unstable...
I guess I better switch back to 1333mhz with EIST on to save power.

Quote:
Originally Posted by Posted: Dec 30, 2008 by Syscev
I only received a noticed about your replies today.
I installed bios v1.35. Turn XMP on and everything else on auto. Windows XP cannot even boot so, I guess that's a fail.
Reply With Quote


  #26  
Old 12-30-2008, 06:30 PM
will160628 will160628 is offline
Registered User
will160628's PC Specs
 
Join Date: Dec 2008
Posts: 10
POST ID # = 382488
will160628 Reputation: 10
Thumbs up Possible fix

I got some new sticks from a different manufacturer as a test and they are not stable at default voltages. However, set the QPI up to +0.18-+0.21 and its rock solid. I've been running 1600Mhz with manual (board won't even find the XMP function on these modules) timings of 8-8-8-24 for almost a week. The only voltages I have changed are the memory (1.65), QPI (+0.20), and ICH (1.60). No need to change the ICH on your end unless you're running raid but on my end it made my array more stable and a bit faster. Now +0.21 seems like a bit much for QPI but I think that the info MSI put in the bios for it holds true. It looks like this board only feeds 1.045v to the QPI with the 920/940 while it feeds the 965 1.225v. I haven't put that to the test with a voltmeter but it definitely seems believable with the voltage I'm having to use. Oh, make sure to turn the power saving feature off for the QPI and memory if you notice any instabilities. I've been trying for 1800Mhz on this new stuff but I just can't quite get the board stable at that speed yet.
Reply With Quote


  #27  
Old 12-30-2008, 09:50 PM
syscev syscev is offline
Registered User
 
Join Date: Nov 2008
Posts: 30
POST ID # = 382517
syscev Reputation: 10
Default Hmmmm....

Quote:
Originally Posted by will160628 View Post
I got some new sticks from a different manufacturer as a test and they are not stable at default voltages. However, set the QPI up to +0.18-+0.21 and its rock solid. I've been running 1600Mhz with manual (board won't even find the XMP function on these modules) timings of 8-8-8-24 for almost a week. The only voltages I have changed are the memory (1.65), QPI (+0.20), and ICH (1.60). No need to change the ICH on your end unless you're running raid but on my end it made my array more stable and a bit faster. Now +0.21 seems like a bit much for QPI but I think that the info MSI put in the bios for it holds true. It looks like this board only feeds 1.045v to the QPI with the 920/940 while it feeds the 965 1.225v. I haven't put that to the test with a voltmeter but it definitely seems believable with the voltage I'm having to use. Oh, make sure to turn the power saving feature off for the QPI and memory if you notice any instabilities. I've been trying for 1800Mhz on this new stuff but I just can't quite get the board stable at that speed yet.
Hmmm.. I could try that.. But I wonder how stable is your stable. I thought I was stable at 1600MHz at first but gaming proved me wrong.

Can you pass 5 passes of Max Stress test using the IntelBurnTest http://www.softpedia.com/get/System/...BurnTest.shtml

Word of warning this thing really really burns up your CPU but the test is very fast compared to all the others.
Reply With Quote


  #28  
Old 12-31-2008, 12:47 PM
will160628 will160628 is offline
Registered User
will160628's PC Specs
 
Join Date: Dec 2008
Posts: 10
POST ID # = 382568
will160628 Reputation: 10
Wink It's stable

I ran Prime 95 for 21 hours on it. I have my processor clocked at 3.2Ghz at 1.3v right now too though. So far I've played about 15-30 hours of Alone in the dark, 10 hours of GTA San Andreas and about 8 hours of Rise of the Argonauts (haven't slept much since last Wensday ). I think it should be stable but I'll try that one out next. I haven't had a sudden reboot since I set these latest voltages. Very happy with these new settings and it amazes me that this thing is actually noticeably faster than my e8400 at 3.6Ghz in my day to day tasks. I'm even considering trying out a virtual machine to let my mom play games on my PC. I just have to find one that actually supports directx with some amount of stability now. Oh and no worries on my cpu temp. I decided to try a few different things with my cooling this go round and it has paid off. My temps haven't passed 62C on any of the cores under any load i throw at it.
Reply With Quote


  #29  
Old 12-31-2008, 08:40 PM
DerekT's Avatar
DerekT DerekT is offline
Registered User
DerekT's PC Specs
 
Join Date: Jul 2005
Location: White Rock, British Columbia
Posts: 10,168
POST ID # = 382624
DerekT Reputation: 10
Default

New hardware means that often you have to be the one that figures the stability out. Let us know if any issues arise.

Quote:
Originally Posted by will160628 View Post
I ran Prime 95 for 21 hours on it. I have my processor clocked at 3.2Ghz at 1.3v right now too though. So far I've played about 15-30 hours of Alone in the dark, 10 hours of GTA San Andreas and about 8 hours of Rise of the Argonauts (haven't slept much since last Wensday ). I think it should be stable but I'll try that one out next. I haven't had a sudden reboot since I set these latest voltages. Very happy with these new settings and it amazes me that this thing is actually noticeably faster than my e8400 at 3.6Ghz in my day to day tasks. I'm even considering trying out a virtual machine to let my mom play games on my PC. I just have to find one that actually supports directx with some amount of stability now. Oh and no worries on my cpu temp. I decided to try a few different things with my cooling this go round and it has paid off. My temps haven't passed 62C on any of the cores under any load i throw at it.
Reply With Quote


  #30  
Old 01-01-2009, 03:41 AM
will160628 will160628 is offline
Registered User
will160628's PC Specs
 
Join Date: Dec 2008
Posts: 10
POST ID # = 382659
will160628 Reputation: 10
Talking Stable

syscev:
IntelBurnTest v1.9 - Utilizes Intel(R) Linpack technology
by AgentGOD
Thanks to Cronos from XS for the memory size equation

Description: Linpack by Intel(R) is an extremely stressful program that will
put even the most powerful X86/X64 CPU in the world at its knees. Load temp
under Linpack will be up to 22*C higher than the competing software Prime95.
This program will make usage of Linpack easier and more practical.

Note: Testing under 64-bit Windows will yield the most accurate result, and is
highly recommended. Please read the included readme file for more information.

Operating Systems supported:
32-bit Windows (X86)
64-bit Windows (X64) [not Itanium]

Currently running OS: 64-bit
Physical memory available: 4719 MB

Turn on error detection (Y/N)? n

Stress levels available:
1 - Maximum stress
2 - Half stress (uses 1/2 of available memory)
3 - Low stress (uses 1/4 of available memory)
4 - Customize stress level

Please select desired stress level> 1
Using maximum available memory to stress test (4719 MB).

Enter the number of times to run the test (5 or more recommended)> 10
----------------------------------------------------
Executing Intel(R) Linpack 64-bit mode...
----------------------------------------------------
Intel(R) LINPACK data

Current date/time: Wed Dec 31 20:13:51 2008

CPU frequency: 3.212 GHz
Number of CPUs: 8
Number of threads: 8

Parameters are set to:

Number of tests : 1
Number of equations to solve (problem size) : 24313
Leading dimension of array : 24313
Number of trials to run : 10
Data alignment value (in Kbytes) : 4

Maximum memory requested that can be used = 434498812, at the size = 24313

============= Timing linear equation system solver =================

Size LDA Align. Time(s) GFlops Residual Residual(norm)
24313 24313 4 280.494 34.1629 5.848528e-010 3.512264e-002
24313 24313 4 280.157 34.2039 5.848528e-010 3.512264e-002
24313 24313 4 280.944 34.1082 5.848528e-010 3.512264e-002
24313 24313 4 283.309 33.8235 5.848528e-010 3.512264e-002
24313 24313 4 278.034 34.4652 5.848528e-010 3.512264e-002
24313 24313 4 280.963 34.1059 5.848528e-010 3.512264e-002
24313 24313 4 279.998 34.2234 5.848528e-010 3.512264e-002
24313 24313 4 279.660 34.2647 5.848528e-010 3.512264e-002
24313 24313 4 277.354 34.5496 5.848528e-010 3.512264e-002
24313 24313 4 281.458 34.0458 5.848528e-010 3.512264e-002

Performance Summary (GFlops)

Size LDA Align. Average Maximal
24313 24313 4 34.1953 34.5496

End of tests

----------------------------------------------------
Testing finished in 3088.289 seconds.
Please check the "Residual(norm)" values and confirm that they are
all equal. If not, your CPU/RAM has failed Linpack testing.
----------------------------------------------------
Please donate if you like this program!
(http://www.ultimate-filez.com/?page=donate)
Execution stopped. Press enter to exit.


I did have some problems with lock-ups if I tried to run anything else with this test and I had to run it at low priority just to keep it from causing a lock-up by its self but the test finally went through once I left it alone. I'm too lazy to turn off HT so temps never went over 48C.

This is all under bios 1.3b3. I haven't noticed a single problem with these beta bios so far other than MSIs auto update wanting to "update" them to version 1.2. I'm actually very happy with the beta drivers right now but I think I'll give the new bios a shot to see if they recognize the xmp function on my new modules. BTW I did try the loosest timings this board will allow with the setting right before red for all voltages and still couldn't get the Corsair sticks stable at 1600 that's why I bought some more modules. They were cheaper but have better timing and an actual return policy so I figured I couldn't go wrong. The gaurantee they would hit 1800Mhz didn't hurt either though .

I wish you luck on 1600mhz. Try upping your QPI to somewhere around +0.25 and see what happens. If its stable drop it till you find the sweet spot. I found the fix on another forum buried somewhere out on google. If you have any questions, I'll be watching this thread for a bit and I'd be happy to help.
Reply With Quote


Reply

Tags
i7

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -4. The time now is 05:11 PM.


Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2019, vBulletin Solutions, Inc.