GA-EP43-UD3L and E7400 FSB Issue

[Mattress Tester]

The setup:


GA-EP43-UD3L, BIOS F3

E7400 Step R0, VID 1.2875
OCZ Vanquisher HSF

2x2GB Corsair DDR2-800 5-5-5-18 1.8v
OCZ 600w StealthXStream PSU


My goal is to hit (9.5 x 400) 3.8GHz with memory at 1:1. So far the best I've done is (9 x 380) 3.6GHz. A 400 FSB seems to be a real problem. I can boot into Windows and run Prime95 for a few minutes at 9 x 400 but increasing CPU and MCH volts hasn't helped. Those setting are listed below. Any tips on how to improve this will be appreciated!


BTW, I not sure if temps are a problem. The core sensors on the CPU never go below 36c and when under load they always report a few degrees lower than the CPU reading in Hardware Monitor (Core 50c vs. Case 55c) which I thought were supposed to be the other way around. So, I've been using the case for reference.

<o:p> </o:p>
MB Intelligent Tweaker (M.I.T.)<o:p></o:p>
Robust Graphics Booster .................: Auto <o:p></o:p>
CPU Clock Ratio .........................: 9 <o:p></o:p>
CPU Frequency ...........................: 3.6<o:p></o:p>
<o:p> </o:p>
Clock Chip Control<o:p></o:p>
Standard Clock Control<o:p></o:p>
CPU Host Clock Control...................: [Enabled] <o:p></o:p>
CPU Host Frequency (Mhz) ................: 400<o:p></o:p>
PCI Express Frequency (Mhz) .............: 100<o:p></o:p>
C.I.A.2 .................................: [Disabled] <o:p></o:p>
<o:p> </o:p>
Advanced Clock Control <o:p></o:p>
CPU Clock Drive..........................: 800mv<o:p></o:p>
PCI Express Clock Drive..................: 900mv<o:p></o:p>
CPU Clock Skew (ps)......................: ops<o:p></o:p>
MCH Clock Skew (ps)......................: ops<o:p></o:p>
<o:p> </o:p>
DRAM Performance Control<o:p></o:p>
Performance Enhance......................: [STANDARD]<o:p></o:p>
(G)MCH Frequency Latch...................: [Auto] <o:p></o:p>
System Memory Multiplier ................: 2.00D<o:p></o:p>
Memory Frequency (Mhz) ..................: 800 <o:p></o:p>
DRAM Timing Selectable ..................: [Manual]<o:p></o:p>
<o:p> </o:p>
Standard Timing Control<o:p></o:p>
CAS Latency Time.........................: 5 <o:p></o:p>
tRCD ....................................: 5 <o:p></o:p>
tRP'.....................................: 5 <o:p></o:p>
tRAS.....................................: 18 <o:p></o:p>
<o:p> </o:p>
Advanced Timing Control<o:p></o:p>
tRRD.....................................: 3 <o:p></o:p>
tWTR.....................................: 3 <o:p></o:p>
tWR......................................: 6 <o:p></o:p>
tRFC.....................................: 42 <o:p></o:p>
tRTP.....................................: 3 <o:p></o:p>
Command Rate (CMD).......................: 2<o:p></o:p>
<o:p> </o:p>
Driving Strength Profiles<o:p></o:p>
Driving Strength ........................: Auto<o:p></o:p>
<o:p> </o:p>
Channel A<o:p></o:p>
<o:p> </o:p>
Static tRead Value.......................: Auto <o:p></o:p>
tRD Phase0 Adjustment....................: Auto<o:p></o:p>
tRD Phase1 Adjustment....................: Auto <o:p></o:p>
tRD Phase2 Adjustment ...................: Auto <o:p></o:p>
tRD Phase3 Adjustment....................: Auto <o:p></o:p>
Trd2rd(Different Rank)...................: Auto <o:p></o:p>
Twr2wr(Different Rank)...................: Auto <o:p></o:p>
Twr2rd(Different Rank)...................: Auto <o:p></o:p>
Trd2wr(Same/Diff Rank)...................: Auto <o:p></o:p>
Dimm1 Clock Skew Control.................: Auto <o:p></o:p>
Dimm2 Clock Skew Control.................: Auto
DDR Write Training.......................: Auto <o:p></o:p>
<o:p> </o:p>

<o:p> </o:p>
Motherboard Voltage Control<o:p></o:p>
Voltage Type.............................: Manual<o:p></o:p>
CPU Vcore.................1.2875v........: 1.36875<o:p></o:p>
<o:p></o:p>
CPU Termination...........1.200v.........: 1.200<o:p></o:p>
CPU Reference.............0.805v.........: Auto <o:p></o:p>
<o:p> </o:p>
MCH/ICH<o:p></o:p>
MCH Core..................1.100V.........: 1.200 <o:p></o:p>
ICH I/O...................1.550V.........: 1.550 <o:p></o:p>
<o:p> </o:p>
DRAM<o:p></o:p>
DRAM Voltage..............1.800V.........: 1.900<o:p></o:p>
<o:p> </o:p>
Advanced Settings<o:p></o:p>
Limit CPUID Max to 3.....................: Disabled<o:p></o:p>
No-execute Memory Protect................: Enabled
C2/C2E State Support.....................: Disabled
C4/C4E State Support.....................: Disabled
CPU enhanced Memory Protect..............: Disabled <o:p></o:p>
CPU Thermal Monitor......................: Enabled<o:p></o:p>
CPU EIST Function........................: Disabled<o:p></o:p>
<o:p> </o:p>

[kick]

Re: GA-EP43-UD3L and E7400 FSB Issue

You have already had the memory running at 800 or more on a multiplier yes?.
Try the 333mhz B multiplier instead of 2:00D

Try turbo performance on the memory.

Try 399 or 401mhz

Try using the xmp setting (to use the memorys epp)with performance on standard and all the timings on auto, ram voltage too.You may/may not have to set the multiplier.

Does the 43 chipset get hot?,if it isnt too bad perhaps a little more mch voltage.
Try the above seperately.
7400 v, I dont know whats needed,but there are a few 7000 series settings posts around.(not specifically for ep43)
Tyremaster was a recent poster ,etc.

Theres no guarantee your hw monitor sensor are reading the right thing. Try a few different utilities.I thought my system temp was reading the northbrige temp and my cpu temp was system temp etc :)
For me the cpu temp is v low at idle, lower than core temps.It shows 9 degrees approx above room temp which is about right .

[Mattress Tester]

Re: GA-EP43-UD3L and E7400 FSB Issue

Just a quick Thank You to Kick!

Your first suggestion to change the multiplier from D to B has allowed me to run Prime95 at 3.6GHz (400 x 9) for three hours and counting. Before that it wouldn't run 10 minutes at 400. I was starting to think I had defective parts. Looks more like 'operator error'.

If this holds up overnight I'll see if I can hit the 400 x 9.5 mark tomorrow and will post the result. Again, thanks for your suggestions.

[Mattress Tester]

Re: GA-EP43-UD3L and E7400 FSB Issue

After success getting stable at 3.6 (400 x 9) I seem to have hit a wall going any further. Other than the divider, no luck with the other suggestions.

- Temps on the cores report the same on CoreTemp, HW Monitor and Everest and top out at in the upper 40's under load. They are just stuck going below 36c.

- As the NB seemed a little warm to the touch I put a 40mm 3,500 RPM fan on it shortly after I assembled the system. Cool to the touch now.

- The P43 doesn't have as many voltage settings as most boards show on this forum. Have tried several combinations at 3.8GHz with no success. As Vcore was increased to 1.4v it started creating other problems (some USB ports stopped working, system got very sluggish and ultimately crashed). Maybe the memory needs further tweaking?

- I do have access to some other memory to try.
.............. OCZ Titanium 2 x 2GB PC2 6400 4-4-4-15 2.2v
.............. Crucial Ballistix 2 x 1GB PC2 6400 4-4-4-12 2.0v

- Currently running with Corsair 2 x 2GB of XMS2 PC2 6400 5-5-5-15 1.8v ($15 AR from Fry's). Passed an hour or two on MemTest.

- My settings are now the same as in the OP except the divider is 2.00B.

Thanks in advance for any help on the quest for 3.8 !

[Mattress Tester]

Re: GA-EP43-UD3L and E7400 FSB Issue

There is a new BIOS out (F4) that shows support removed for ITE8718 KX, which was a feature added to BIOS F2. Should I try the new BIOS? Does the removed feature have anything to do with issues overclocking this board?