Hardware Canucks

Hardware Canucks (http://www.hardwarecanucks.com/forum/)
-   Troubleshooting (http://www.hardwarecanucks.com/forum/troubleshooting/)
-   -   I7 920 with 12GB 1600mhz ddr3, HELP with system stability issues!!! (http://www.hardwarecanucks.com/forum/troubleshooting/23767-i7-920-12gb-1600mhz-ddr3-help-system-stability-issues.html)

HH89 September 30, 2009 03:06 AM

I7 920 with 12GB 1600mhz ddr3, HELP with system stability issues!!!
 
Hey guys,

First time posting here, what a great community y'all have here. =)

Anyways, I recently purchased a prebuilt computer with the current components:

Asus P6T (regular) Motherboard.
Core i7 920 2.67ghz processor.
12gb (two sets of triple channel 3x2gb sticks) OCZ ddr3 1600mhz low voltage ram (1.65v).
2x Asus EAH4850 Video cards
Coolermaster V8 CPU Cooler
Antec 850W PSU
1TB WD harddrive which I have the OS installed on.
2TB WD harddrive for storage.
There are 5 fans installed (not counting the 2 GPU fans or the PSU Fan - a total of 8 fans if I were to count those in as well)
Case is a Thermaltake Armor+ VH600.

I've never overclocked in the past so please bear with me here.

From what I've read so far, the core i7 920 can only run 1333mhz/1600mhz+ ram if I manually tweak (overclock) the cpu clock speeds (is this refered to as the "blck"?) so that the CPU clock speed is twice as much as the Dram Frequency??? (correct me if im wrong)

So in otherwords, to run the ram I have at its rated 1600mhz, 7-7-7-24 timings I have to overclock the CPU to at least 3.2ghz? to run stable? correct? (again, plz correct me if im wrong).

Anyways, I checked the BIOS settings and the people who built my computer had only set the dram frequency manually to 1600mhz at 7-7-7-24 timings with a Dram Bus Voltage of 1.66V and left everything else at "Auto".

Anyways, heres the problem:

This computer I recently bought has been running into stability issues. I would be doing some database work on it, importing information into a database, and it would bsod/restart on me once every 3-4 hrs. It also BSOD'd on me several times whenever I tried running windows update after a fresh reformat. But I was able to get past this problem by installing all the chipset drivers first before running windows update on a fresh reformat (perhaps it was just a temporary solution?).

So, I decided to run stress/stability tests on it to test the system's stability.

Running Prime95 64bit (blend test) within 2 minutes errors would show up, then within 5 minutes my computer would just restart itself. I ran this about 4-5 times each with the same results (I would get maybe 2-4 errors during the test before the computer would restart itself). Keep in mind I have not done any overclocking at all for the processor, nor have I changed any of the voltage settings (everything is still set to auto), the only thing that has been set higher then default settings are the ram settings, which again the builders of the computer had to set at 1600mhz, 7-7-7-24 timings manually. From what I understand, a Core i7 920 is supposed to default run the ram at 1066mhz settings, correct?

So anyways, the Prime95 test was a huge fail.

Is this a result of bad hardware? Perhaps faulty CPU or memory? Or am I supposed to key in the clock speeds/voltage settings manually for everything if the ram settings are already keyed in manually.. in order to run this system stable?

Also, are the computer restarts while running Prime95 due to my CPU overheating? Which is why I can't finish the test? I was observing the temps for the CPU using PC Probe during those tests and it would go up to about the 63-68C range before the comp would restart. Does the P6T mobo have some sort of CPU temperature threshold where it will restart the computer if it detects the CPU temperature going over 68C for example? Or are those temperatures normal while running this test? For what its worth, my idle cpu/mobo temperatures are about 42/39 respectively.

So the next test I ran was Super PI. Within 2-3 min and error would come up. I ran it twice, the first time it would just encounter a windows error and force me to shutdown the program after I closed the error box. The 2nd time a "rounding error" came up.

Third, I ran Intel Burn Test. I just ran the test at "standard settings" and it ran to completion successfully. I have yet to try this test at a higher setting. Will probably try this test again sometime later today.



Based on this computer's specs and these results, does anyone know what could possibly be causing these stability issues?

I am currently running Memtest86+ V4.00 on it. So far it has run for 3 hours (2 passes) with 0 errors. Will leave this on over night then post the results here. However, to be honest, I'm not sure if i'm using it correctly (and by that I mean I'm not sure if i'm running a thorough enough test using it?). All I did was just boot it off the CD and let it do its thing, I did not change anything in the configurations. Am I supposed too?
Anyone with experience in this, please let me know. =)

Anyways, any suggestions as to what I should do to try to solve this problem? I figure I should try testing things out on my own before sending this back to their testing department, since its a huge pain in the *** to send it back.

From what I've read, I'm supposed to be able to run Prime95 for over 24hrs to be sure of a stable system. Right now I can't even run it for more then 5 minutes without the test failing or the computer restarting.

Again, I did not overclock and to be honest I know next to nothing about OCing (other then what I've read on several OCing forums), so if possible, would like to avoid OCing the processor and possibly voiding the warranty??

Any help is appreciated. Thanks in advance.


Kind regards,
- Carl




__________________________________________________ _________


Cliff notes:

core i7 920 processor
12gb ddr3 1600mhz ram.
2x asus eah4850

Current bios configuration has the ram set at stock settings of 1600mhz, 7-7-7-24 timings and 1.66volts.
Everything else in the bios set to auto.

Prime 95 Stability test fails within 5 minutes with multiple errors coming up, computer restarts itself within 5 mins of running Prime 95.

Super Pi also errors within 5 mins of having it running.

Testing with Memtest86+ on default settings, so far its ran for 14hrs, 7 passes, with 0 errors. (any way to do a more thorough memtest86+ scan?).

Based on my situation, how would I go about diagnosing and fixing this problem?
Any solutions?

(I have the exact same problem happening on 2 computers which I bought with exact same specs, only difference is one has a 900watt PSU, the other a 850watt PSU).


The return process takes a really long time, so if possible if I could diagnose the problem and fix it here without taking it back that would be preferable. The solution may be as simple as manually keying in the bios settings (clock speeds/voltage settings?) perhaps, which is what im wondering?

martin_metal_88 September 30, 2009 04:44 AM

OCZ memory can be an issue. Most of the time they need to be set manually in bios because they suck more voltage that they are suppose to. You can ask on OCZ forum with your ram model and your board and they will reply the exact setting you need to get for the ram to run great. The core I7 920 temp limit is supose to be about 65 to 70. It's normal that the computer would reboot at like 68. But it is not that a not OC cpu would get at this temp. Core I7 920 full load at stock speed get about 45 to 50 degree in a normal time. Take a look at your cooler sitting and thermal paste.

Ask for more.

bojangles September 30, 2009 04:55 AM

Quote:

So in otherwords, to run the ram I have at its rated 1600mhz, 7-7-7-24 timings I have to overclock the CPU to at least 3.2ghz? to run stable? correct? (again, plz correct me if im wrong).
No. Run the memory multiplier at 2:12 or 1:6 and it will automatically go to 1600 MHz. Manually set your timings as well and set RAM voltage to 1.65 V.

Quote:

Is this a result of bad hardware? Perhaps faulty CPU or memory? Or am I supposed to key in the clock speeds/voltage settings manually for everything if the ram settings are already keyed in manually.. in order to run this system stable?
Sounds more like a memory issue to me. One of your sticks could be dying and causing errors. Your temps are fine on the CPU. I run mine at 80C in the summer with no problems. The result of your P95 and SuperPI tests seem to point to RAM issues. If you really want to find out, set the CPU multiplier to something lower than 20, like 15 say to see if it's a frequency problem. Does the RAM work at specified speed? Try a lower speed there too (lower memory multiplier), but don't run them lower at the same time. Also try taking a kit out and see if 6GB works (typically you won't need 12GB of RAM lol :haha:). Also remember that the uncore frequenct has to be DOUBLE the memory frequency or more.

You may be also running into stability issues because all your RAM slots are being used. Sometimes it's just too hard on the CPU and can cause some problems. Perhaps try raising the Uncore voltage a bit and see if that helps your memory. There may be too much power getting sucked from your RAM and it's causing high voltage sag. Who knows.

Intel Burn in Test is crap IMHO. It stresses the CPU, but nothing else. Stick with Prime95, SuperPI 32M (on all 4/8 cores, depending if you run with HT on), and Linx. Run the blend test in Prime95, as it tests lots of CPU and RAM.

Good luck. HTH.:thumb:

Arinoth September 30, 2009 04:57 AM

Quote:

Originally Posted by martin_metal_88 (Post 263536)
OCZ memory can be an issue. Most of the time they need to be set manually in bios because they suck more voltage that they are suppose to. You can ask on OCZ forum with your ram model and your board and they will reply the exact setting you need to get for the ram to run great. The core I7 920 temp limit is supose to be about 65 to 70. It's normal that the computer would reboot at like 68. But it is not that a not OC cpu would get at this temp. Core I7 920 full load at stock speed get about 45 to 50 degree in a normal time. Take a look at your cooler sitting and thermal paste.

Ask for more.

What the hell are you talking about? the core i7 920 temp limit is about 90C and its not at all normal for the cpu to be rebooting at 68. Normal is it normal at stock speeds to be full load of 45-50C (assuming we're using the stock HSF). Myself and others run theirs on air overclocked anywhere from 65-80C stable. Stock full load with stock fan will run you about 70-75C, potentially cooler if you have a cool room. If you don't know your facts on a cpu don't give assumptions

gingerbee September 30, 2009 05:03 AM

would it bsod at 68c core temp no ( 67c for the ihs ) tm monitoring starts throttling at like 100c core temp. I run mine all day long at above 70c core

martin_metal_88 September 30, 2009 05:25 AM

Quote:

Originally Posted by gingerbee (Post 263543)
would it bsod at 68c core temp no ( 67c for the ihs ) tm monitoring starts throttling at like 100c core temp. I run mine all day long at above 70c core

Soryr guys mix it with the recomendation of AMD 920 XD ( suppoe to be around 62 according to AMD ). Morning have a wierd effect on me:biggrin:.

Ask for more.

bojangles September 30, 2009 07:30 AM

Quote:

Originally Posted by martin_metal_88 (Post 263555)
Soryr guys mix it with the recomendation of AMD 920 XD ( suppoe to be around 62 according to AMD ). Morning have a wierd effect on me:biggrin:.

Ask for more.

What???? AMD CPUs can run hot too. Where do you get this info from?

EDIT: Ohhhh I see. That doesn't do anything though if you disable it :haha:.

Squeetard September 30, 2009 07:35 AM

Dude try your ram at 8-8-8-4 and see if it is stable. I had to set my ram voltage to 1.7 on my ocz, and it still drops to 1.64 under load. set your cpu pll voltage to 1.3 also.

bojangles September 30, 2009 08:03 AM

Quote:

Originally Posted by Squeetard (Post 263598)
Dude try your ram at 8-8-8-24 and see if it is stable. I had to set my ram voltage to 1.7 on my ocz, and it still drops to 1.64 under load. set your cpu pll voltage to 1.3 also.

fixed.

Squeetard September 30, 2009 08:25 AM

yah, sorry. The 0 and the 2 on this G15 no longer work. I had to assign them to the macro keys, forgot.


All times are GMT -7. The time now is 09:13 PM.