F2-8500CL5D-4GBPK + GIGABYTE GA-EP45-UD3P / R BIOS settings

[Soulone]

Q9550 2.83 ghz UD3R rev1.1 F12 bios
Moved sticks to slots2,4 Even though I have the F12 BIOS
Ran Memtest 4.0 for 3hrs.39min. no errors
And 20+20 intel burn tests no errors.Seems like I have been here before and it failed memtest in 2hours, except sticks were in1,3 slots and I think I had xmp on auto.
Wonder about tightening timings maybe, since mem is only running at 912?



MB Intelligent Tweaker(M.I.T.)
Robust Graphics Booster ...[AUTO]
CPU Clock Ratio ...........8X
Fine CPU Clock Ratio.......+0.5
CPU Frequency .............3.23 ghz (380x8.5)

Clock Chip Control
>>>Standard Clock Control
CPU Host Clock Control.......[Enabled]
CPU Host Frequency (Mhz) ....380
PCI Express Frequency (Mhz)..AUTO

C.I.A.2 .....................[Disabled]

Advanced Clock Control [Press Enter]


DRAM Performance Control
Performance Enhance..............[STANDARD]
Extreme Memory Profile (X.M.P.)..[Disabled]
(G)MCH Frequency Latch...........[Auto]
System Memory Multiplier ........[2.40B]
Memory Frequency (Mhz) ..........1066 912
DRAM Timing Selectable ..........[Manual]

Standard Timing Control
CAS Latency Time...5 [5]
tRCD ..............7 [5]
tRP'...............7 [5]
tRAS...............20 [15]

Advanced Timing Control
tRRD...............4 [AUTO]
tWTR...............4 [AUTO]
tWR................8 [AUTO]
tRFC...............68 [60]
tRTP...............4 [AUTO]
Command Rate (CMD).0 [2]






Motherboard Voltage Control
Voltage Type
>>> CPU
Load-Line Calibration [DISABLED]
CPU Vcore...............1.3000v..[1.28125v..]
CPU Termination?........1.200v....[AUTO]
CPU PLL????.............1.500v....[AUTO]
CPU Referen............:0.760v....[AUTO]

>>>MCH/ICH
MCH Core.........1.100v...[AUTO]
MCH Reference....0.760v...[AUTO]
MCH/DRAM Ref.....0.900v...[AUTO]
ICH I/O..........1.500v...[AUTO]
ICH Core.........1.100v...[AUTO]

>>> DRAM
DRAM Voltage ........1.800v...[2.100v]
DRAM Termination.....0.900v...[AUTO]
Channel A Reference..0.900v...[AUTO]
Channel B Reference..0.900v...[AUTO]

[Soulone]

Oh, Just seen that: tRFC is set to
tRFC 68 [AUTO]....lol I copied and pasted this and missed that setting...lol

[Soulone]

Running good so far with above settings and also was able to set timings to 4-5-5-15 , Passed memtest with these timings...Wouldn`t do 4-4-4-12, haven`t tried 4-5-4-14 yet

[GSKILL TECH]

DDR2-912 should be able to do 4-5-4-14, glad to hear everything is shaping up.

Thank you
GSKILL TECH

[Soulone]

Yes it runs at 4-5-4-14 just fine.. Memtest for 3hrs.45min. no errors! Don`t know if the slots2,4 made any difference, I think having the xmp set to auto instead of disabled might have been the problem there and not the 1,3 slots. Will it hurt anything to leave the ram in slots2,4 ?? And from what I`ve read in the forums, it sounds like running this ram at 912 with 4-5-4-14 timings will actually be faster than 1066 at 5-5-5-15 timings...correct?

[GSKILL TECH]

Surely is possible that XMP was causing the problem. We had the same problem with our tests. Slots 2 and 4 should be no problem.

You can test memory bandwidth to see which is actually faster. Everest Ultimate is a nice program that you can use for this purpose.

Thank you
GSKILL TECH

[Soulone]

Thanks for all the great support!! Everything seems to be running smoothly now.

[dboon]

still have problems? continued from April 15

I'm slow to come back to this system because it's so frustrating. OK, so I'm trying again. I plan to reinstall Win 7, but first I want to make sure the memory works. Using Memtest86+ v4.10, I got three failures in two hours:
one at address 000401 48684
and two at 000401 d8e84

I don't want to reinstall Win 7 with failing memory ? that will just give me another corrupted Win 7 like I have now.

My BIOS settings are according to my reading of the recommendations at this thread, which are:

BIOS F12
Motherboard Intelligent Tweaker
CPU Clock Ratio 8 (2.66 CPU, underclocked)
CPU Host Clock Control Enabled
CPU Host Frequency 333
Performance Enhancer Standard
Extreme Memory Profile Disable
System Mermory Multiplier (SPD) 3.20B
Dram Timing Selectable Manual
CAS (tCL) Column Address Select 5
tRCD RAS to CAS delay 5
tRP Row Precharge 5
tRAS min RAS active 15
Advanced Timing Control
tRRD Row to Row Delay 4
tWTR Write To Read delay 4
tWR Write Recovery 8
tRFC Row Refresh Cycle 80
tRTP Read To Precharage 4
CPC Command Rate 2
Load Line Calibration Disable
MCH Core Voltage 1.40
Dram Voltage 2.10

As I said in my earlier post, I'm more interested in stability than speed. Please advise on what to try next.

Here's my system:
OS: Windows 7 pro, 64 bit
Mobo: GA-EP45-UD3R rev 1.1
BIOS F12, using settings recommend for F12 in the first message in this thread.
Memory: F2-8500CL5D-4GBPK ? 8 gigs
CPU: Intel E7500
Graphics: nVidia N958 GE Force 9500

...now, after 2 hours, 40 minutes, memtest is up to 17 errors

Thanks.

- dboon

[GSKILL TECH]

tRFC 68
MCH Core Voltage 1.20V

Test each module to see if one may be defective causing the problems.

Thank you
GSKILL TECH

[dboon]

...still having problems

GSkill Tech,

Thanks for the suggestion. I did some more runs with memtest86+ v4.10.

First, let me say that I am not mixing up slots and modules. I numbered 4 envelopes, 1 to 4. I put the module from slot 1 in envelope 1 when not used, and back in slot 1 when used, etc. Of course, 1 & 3 and 2 & 4 are matched modules.

I changed tRFC and MCH voltage according to your recommendation, then ran memtest against each module individually ? module 1 in slot 1, etc. for three passes, which took about an hour and a quarter each. No errors.

Then I ran memtest against modules 1 and 3 in slots 1 and 3. Three passes took about 2 and a half hours and gave no errors.

I am now running memtest against modules 2 and 4 in slots 2 and 4. I am getting the same errors that I got yesterday when testing all 8 gig.

The errors all have this in common:
- The failing address starts with 0040.
- The "Err-bits" are always 00010000
- The errors occur on Test #7, Mod 20 Random bits.

Please Advise.

Thanks.

- dboon

[GSKILL TECH]

Test modules 2 and 4 individually to see which may be defective.

Thank you
GSKILL TECH

[dboon]

...still having problems

Already did that. The modules are good individually, but fail when together. Please re-read my earlier post.

[GSKILL TECH]

Have you tested modules 2 and 4 in slots 1 and 3?

Thank you
GSKILL TECH

[dboon]

? closing in on the problem

GSKILL TECH,

Thanks for your response.. Testing Modules 2 & 4 in slots 1 & 3 is a good idea. The result is memory failure, of the same type as noted before.

To summarize: I have four memory modules that I am calling 1, 2, 3, and 4 after their original slot positions. If I test one module at a time, they all pass Memtest86+ v4.10. Testing two modules at a time, modules 1 & 3 pass in slots 1 & 3. They also pass if moved to slots 2 & 4. Modules 2 & 4 fail in slots 2 & 4. They also fail if moved to slots 1 & 3. The memory failures, then, are associated with modules 2 & 4, not the motherboard slots. Also, the failures are associated with the pairing of 2 & 4, not the individual modules (which pass when tested individually.) Did I state that clearly?

Please advise, and thanks for your patience.

- dboon

[GSKILL TECH]

Please send them in for new replacements and that should solve the problem.

http://www.gskill.com/rma.php

Thank you
GSKILL TECH