X48 Rampage Formula Preview.

**leo27** · 12-25-2010, 08:54 PM

I've also been doing some tweaking and it seems I hit a FSB wall at 432Mhz, at 433Mhz the boot process gets stuck at DETDRAM even though I've tested the ram to work at 1200Mhz at 1.80v.

I probably have a FSB Hole around the 433-440Mhz or so mark but I haven't really tested that out.

the Settings I used are:

Code:

*Extreme Tweaker* 
Ai Overclock Tuner               [Manual] 
OC From CPU Level Up             [Auto] 
CPU Ratio Setting                [08.0] 
FSB Strap to North Bridge        [400] 
FSB Frequency                    [432] 
PCIE Frequency                   [103] 

DRAM Frequency                   [DDR2-1152MHz] 
DRAM Command Rate                [2N] 
DRAM CMD Skew on Channal A       [Greyed Out] 
DRAM CMD Skew on Channal B       [Greyed Out] 
DRAM CLK Skew on Channal A       [Auto] 
DRAM CLK Skew on Channal B       [Auto] 

DRAM Timing Control              [Manual] 
 1st Information 
  CAS# Latency                   [5] 	
  RAS# to CAS# Delay             [5] 	
  RAS# PRE Time                  [5] 	
  RAS# ACT Time                  [15] 	
  RAS# to RAS# Delay             [1] 
  REF Cycle Time                 [Auto (52)] 
  WRITE Recovery Time            [2] 	
  READ to PRE Time               [1] 
 2nd Information 
  READ to WRITE Delay (S/D)      [8] 
  WRITE to READ Delay (S)        [3] 
  WRITE to READ Delay (D)        [4] 
  READ to READ Delay (S)         [4] 
  READ to READ Delay (D)         [5] 
  WRITE to WRITE Delay (S)       [4] 
  WRITE to WRITE Delay (D)       [5] 
 3rd Information 
  WRITE to PRE Delay             [13]	
  READ to PRE Delay              [4] 	
  PRE to PRE Delay               [1] 	
  ALL PRE to ACT Delay           [4] 	
  ALL PRE to REF Delay           [4] 
  Refresh Period		 [16120T]

DRAM Static Read Control         [Disabled] 
Ai Clock Twister                 [Lighter] 
Ai Transaction Booster           [06] 		
Pull-in of CHA PH1	         [Disabled]	
Pull-in of CHA PH2	         [Disabled]	
Pull-in of CHA PH3	         [Disabled]	
Pull-in of CHA PH4	         [Disabled]	
Pull-in of CHB PH1	         [Disabled]	
Pull-in of CHB PH2	         [Disabled]	
Pull-in of CHB PH3	         [Disabled]	
Pull-in of CHB PH4	         [Disabled]	
CPU Voltage                      [1.45000v] 
CPU PLL Voltage                  [1.50v] 
North Bridge Voltage             [1.25v] 
DRAM Voltage                     [1.80v] 
FSB Termination Voltage          [1.28v] 
South Bridge Voltage             [1.05v] 
SB 1,5V Voltage                  [1.50v] 
Loadline Calibration             [Enabled] 
CPU GTL Voltage Reference        [0.67x] 
NB GTL Voltage Reference         [0.67x] 
DRAM Controller Voltage REF      [AUTO] 
DRAM Channel A Voltage REF       [AUTO] 
DRAM Channel B Voltage REF       [AUTO] 

CPU LED Selection                [CPU Voltage] 
NB LED Selection                 [North Bridge Voltage] 
SB LED Selection                 [South Bridge Voltage] 
Voltiminder LED                  [Enabled] 

CPU Spread Spectrum              [Disabled] 
PCIE Spred Spectrum              [Disabled] 
Cpu Clock Skew		         [Delay 200ps]
NB Clock Skew		         [Delay 100ps]

*CPU Settings* 

CPU Ratio Settings	         [08.0]
C1E Support	                 [Enabled]
CPU TM Function	                 [Enabled]
Vanderpool Technology            [Disabled]
Execute Disabled Bit	         [Enabled]
Max CPUID Value Limit            [Disabled]

But this fails P95+UT2004 Blend very quickly.
As you can see memory is not the issue, the problem is almost certainly FSB Termination voltage however LinX failed at voltages from 1.30-1.36v (around 1.36-1.42v Real)

.

Any suggestions??

**spy2520** · 12-26-2010, 05:30 AM

I can run linx all day at 480. Haven't mastered anything above that though. Although I noticed it relies on temps. If the room goes up a few degress it'll fail randomly. If cpu temps stay closer to 60*c it runs fine.

**Alien Grey** · 12-26-2010, 07:21 AM

Originally Posted by leo27

I've also been doing some tweaking and it seems I hit a FSB wall at 432Mhz, at 433Mhz the boot process gets stuck at DETDRAM even though I've tested the ram to work at 1200Mhz at 1.80v.

I probably have a FSB Hole around the 433-440Mhz or so mark but I haven't really tested that out.

But this fails P95+UT2004 Blend very quickly.
As you can see memory is not the issue, the problem is almost certainly FSB Termination voltage however LinX failed at voltages from 1.30-1.36v (around 1.36-1.42v Real)

.

Any suggestions??

I think that you created the FSB hole yourself by tightening the timings.

Ai Clock Twister Lighter works very well but it doesn't like tight timings. That doesn't really matter because we can reach higher FSB to compensate the loss in performance and that's why I'm using the Rampage Extreme timings that are more relaxed.

When you are almost certain that it's the FSB Termination Voltage that makes it fail try with CPU Voltage Reference 0.63X or 0.65X and NB Voltage Reference 0.63X and higher FSB Termination Voltage.

CPU Clock Skew Delay 300ps and NB Clock Skew Delay 200ps should work very well.

**Alien Grey** · 12-26-2010, 10:58 AM

My new 24/7 Overclock.

Running Prime95 Custom using 3200Mb Memory with Wine in Ubuntu at 4.22GHz and DDR 1250MHz with tRD 6 and tREF 16120T.

It should be stable enough I think to run BOINC\seti@home in Ubuntu.

BIOS Settings:

Code:

*Extreme Tweaker* 
Ai Overclock Tuner               [Manual] 
OC From CPU Level Up             [Auto] 
CPU Ratio Setting                [09.0] 
FSB Strap to North Bridge        [400] 
FSB Frequency                    [469] 
PCIE Frequency                   [100] 

DRAM Frequency                   [DDR2-1250MHz] 
DRAM Command Rate                [2N] 
DRAM CMD Skew on Channal A       [Greyed Out] 
DRAM CMD Skew on Channal B       [Greyed Out] 
DRAM CLK Skew on Channal A       [Auto] 
DRAM CLK Skew on Channal B       [Auto] 

DRAM Timing Control              [Manual] 
 1st Information 
  CAS# Latency                   [5] 	
  RAS# to CAS# Delay             [5] 	
  RAS# PRE Time                  [5] 	
  RAS# ACT Time                  [15] 	
  RAS# to RAS# Delay             [5] 
  REF Cycle Time                 [Auto (52)] 
  WRITE Recovery Time            [12] 	
  READ to PRE Time               [6] 
 2nd Information 
  READ to WRITE Delay (S/D)      [8] 
  WRITE to READ Delay (S)        [6] 
  WRITE to READ Delay (D)        [6] 
  READ to READ Delay (S)         [4] 
  READ to READ Delay (D)         [7] 
  WRITE to WRITE Delay (S)       [4] 
  WRITE to WRITE Delay (D)       [7] 
 3rd Information 
  WRITE to PRE Delay             [23]	
  READ to PRE Delay              [8] 	
  PRE to PRE Delay               [1] 	
  ALL PRE to ACT Delay           [6] 	
  ALL PRE to REF Delay           [6] 
  Refresh Period		 [16120T]

DRAM Static Read Control         [Disabled] 
Ai Clock Twister                 [Lighter] 
Ai Transaction Booster           [06] 		
Pull-in of CHA PH1	         [Disabled]	
Pull-in of CHA PH2	         [Disabled]	
Pull-in of CHA PH3	         [Disabled]	
Pull-in of CHA PH4	         [Disabled]	
Pull-in of CHB PH1	         [Disabled]	
Pull-in of CHB PH2	         [Disabled]	
Pull-in of CHB PH3	         [Disabled]	
Pull-in of CHB PH4	         [Disabled]	
CPU Voltage                      [1.36250v] 
CPU PLL Voltage                  [1.50v] 
North Bridge Voltage             [1.39v] 
DRAM Voltage                     [1.92v] 
FSB Termination Voltage          [1.30v] 
South Bridge Voltage             [1.05v] 
SB 1,5V Voltage                  [1.50v] 
Loadline Calibration             [Disabled] 
CPU GTL Voltage Reference        [0.63x] 
NB GTL Voltage Reference         [0.67x] 
DRAM Controller Voltage REF      [AUTO] 
DRAM Channel A Voltage REF       [AUTO] 
DRAM Channel B Voltage REF       [AUTO] 

CPU LED Selection                [CPU Voltage] 
NB LED Selection                 [North Bridge Voltage] 
SB LED Selection                 [South Bridge Voltage] 
Voltiminder LED                  [Enabled] 

CPU Spread Spectrum              [Disabled] 
PCIE Spred Spectrum              [Disabled] 
Cpu Clock Skew		         [Delay 300ps]
NB Clock Skew		         [Delay 200ps]

*CPU Settings* 

CPU Ratio Settings	         [09.0]
C1E Support	                 [Disabled]
CPU TM Function	                 [Enabled]
Vanderpool Technology            [Disabled]
Execute Disabled Bit	         [Enabled]
Max CPUID Value Limit            [Disabled]

Relaxing the timings with using the Rampage Extreme Memory timings makes it 24/7 stable with CPU Clock Skew Delay 300ps and NB Clock Skew Delay 200ps.

It isn't necessary anymore to make a 50ps difference between DRAM Channel A and B with the more relaxed timings and it's working fine with Auto settings.

With the Rampage Extreme Memory timings it's stable with less CPU Voltage and NB Voltage with CPU Clock Skew Delay 300ps and NB Clock Skew Delay 200ps.

Thanks a lot SoLoR for sharing this fantastic MOD BIOS. You made the impossible possible.

**greg.m** · 12-26-2010, 11:10 AM

Originally Posted by Alien Grey

Thanks a lot SoLoR for sharing this fantastic MOD BIOS. You made the impossible possible.

And which bios are we talking about? The last one with the latest rom?

**Alien Grey** · 12-26-2010, 12:20 PM

Originally Posted by greg.m

And which bios are we talking about? The last one with the latest rom?

Yes, the last one. Or is there already a new one?

**greg.m** · 12-26-2010, 02:59 PM

Originally Posted by SoLoR

BTW yesterday i updated 1002bMOD.rom again. With Intel OROM 10.1.0.1008.

Originally Posted by Alien Grey

Yes, the last one. Or is there already a new one?

I mean the one above ^^ Thats the last.

**leo27** · 12-26-2010, 07:19 PM

Originally Posted by Alien Grey

I think that you created the FSB hole yourself by tightening the timings.

Ai Clock Twister Lighter works very well but it doesn't like tight timings. That doesn't really matter because we can reach higher FSB to compensate the loss in performance and that's why I'm using the Rampage Extreme timings that are more relaxed.

When you are almost certain that it's the FSB Termination Voltage that makes it fail try with CPU Voltage Reference 0.63X or 0.65X and NB Voltage Reference 0.63X and higher FSB Termination Voltage.

CPU Clock Skew Delay 300ps and NB Clock Skew Delay 200ps should work very well.

When i was trying different FSB's I used the auto timings which should have been good since I'm only talking about trying for the 438 level to try and reach 3500Mhz

.

I'll have to have another go at it when I get a cool day (its summer here).

Does anyone know roughly what FSB termination voltage I should be looking at for the 432-438Mhz mark because I'm already at 1.424v Real for 425Mhzx8.0??

BTW Nice work A-Grey

**Alien Grey** · 12-27-2010, 03:25 AM

Originally Posted by greg.m

I mean the one above ^^ Thats the last.

Yes, the one from December 10.

**SoLoR** · 12-27-2010, 05:59 AM

Originally Posted by Alien Grey

Yes, the one from December 10.

"next" one will be dated most likely so you will know when/if its new

**Alien Grey** · 12-27-2010, 12:12 PM

The MOD BIOS 1002 and the ASUS Rampage Extreme timings make my board shine again like in my Q9450 days.

Now LinX fails only with my Q9650 because of to low CPU Voltage like it did with my Q9450. It doesn't matter anymore that it isn't LinX stable. Prime95 stable is enough and to be LinX stable I would need more CPU Voltage that isn't necessary to run 24/7 stable.

I've got to use NB Voltage 1.39V and FSB Termination Voltage 1.30V in the BIOS to run it stable in Windows 7 but I didn't touch any other voltage and hope to have it Prime95 stable running for a few hours in Windows 7 at 4.22GHz and DDR 1250MHz with tRD 6 and tREF 16120T.

**Alien Grey** · 12-27-2010, 01:57 PM

It's Prime95 Blend Custom 3200MB stable in Windows 7 Ultimate X64.

All I had to do is up the NB Voltage to 1.39V and the FSB Termination Voltage to 1.30V in the BIOS.

**leo27** · 12-27-2010, 04:25 PM

Looks Good A-Grey. Although if your going to fold or use BOINC I'd give that a run through just to make sure it doesn't fail from the low vCore.
If that works fine then I'd say your good since no game is going to stress the CPU as much as a folding program.

BTW How do you know that LinX is only failing due to vCPU is it just from the fact it passes P95??

**OLDcomer** · 12-27-2010, 10:47 PM

If you pass Prime95 x64 small and blend and LinX fails then you've better improve your cooling.

**Korgan** · 12-27-2010, 11:27 PM

Hi!

First of all I would like to thank Solor for the excellent bios and Alien Grey for the settings!

I have tried Alien Greys last settings and it works great. I have to set Ai Transaction Booster [06] to [08]. Otherwise it does not boot? I also have to raise North Bridge Voltage to 1.77v. I have mounted a 60mm fan ontop the North Bridge and that keeps the NB below 60c.

Here is some benches using 1.85v on NB.

Any hints to break 500Mhz FSB?

**Alien Grey** · 12-28-2010, 12:10 AM

Originally Posted by leo27

Looks Good A-Grey. Although if your going to fold or use BOINC I'd give that a run through just to make sure it doesn't fail from the low vCore.
If that works fine then I'd say your good since no game is going to stress the CPU as much as a folding program.

BTW How do you know that LinX is only failing due to vCPU is it just from the fact it passes P95??

Because I've tested it. You have to know that I can't use Load Line Calibration so nothing is going to take as much CPU Voltage as Prime95 so the CPU Voltage is never going to be to low.

Remember that at 9 X 450MHz is was LinX stable but couldn't last 2 hours of Prime95 Blend without a BSOD.

I'm going to do some BOINC because there are a lot of packages waiting to get processed. I'll soon know if it's good enough to run BOINC.

**Alien Grey** · 12-28-2010, 12:16 AM

Originally Posted by OLDcomer

If you pass Prime95 x64 small and blend and LinX fails then you've better improve your cooling.

I don't think I've got a cooling problem.

Could you post a screenshot showing the Chipset information from your Giga board? I would like to see the memory timings.

**OLDcomer** · 12-28-2010, 02:46 AM

Originally Posted by Alien Grey

I don't think I've got a cooling problem.

Could you post a screenshot showing the Chipset information from your Giga board? I would like to see the memory timings.

Make a choice and I'll tell you what everest shows:

1. 500fsb

2. 512 fsb

3. 523fsb

I think I've also got 523 fsb picture but with PL9 and strap 400

**OLDcomer** · 12-28-2010, 02:55 AM

Originally Posted by Korgan

Hi!

First of all I would like to thank Solor for the excellent bios and Alien Grey for the settings!

I have tried Alien Greys last settings and it works great. I have to set Ai Transaction Booster [06] to [08]. Otherwise it does not boot? I also have to raise North Bridge Voltage to 1.77v. I have mounted a 60mm fan ontop the North Bridge and that keeps the NB below 60c.

Here is some benches using 1.85v on NB...

Any hints to break 500Mhz FSB?

Your North Bridge Voltage is too high for such fsb speeds.
Check this out at North Bridge Voltage = 1.78V and FSB Termination Voltage = 1.6V:
http://hwbot.org/community/submissio...ghz_9sec_984ms

**Alien Grey** · 12-28-2010, 03:59 AM

Originally Posted by OLDcomer

Make a choice and I'll tell you what everest shows:

Let's start with FSB 500MHz.

@Korgan

FSB 500MHz is next on my list and I hope to have it stable for at least 20 passes in LinX.

I guess I was wrong when I said that LinX failed because of to low CPU Voltage.

But I was right when I said that now she's as stable at FSB 469MHz as she was at FSB 450MHz with any other BIOS.

It isn't the CPU Voltage that was the problem but the CPU Voltage Reference and going back to 0.63X gives me this:

**OLDcomer** · 12-28-2010, 11:52 AM

Originally Posted by Alien Grey

Let's start with FSB 500MHz.
...

**Alien Grey** · 12-28-2010, 12:53 PM

Originally Posted by OLDcomer

Thanks. It's going to be very useful.

**Alien Grey** · 12-29-2010, 05:27 AM

@OLDcomer

There's one timing that I can't find on the EVEREST screenshot. The timing Write Recovery Time. Maybe you can find that one in the BIOS.

MCH Read ODT is also different and can't be changed in the BIOS.

ASUS -> 8T
Gigabyte -> 11T

Edit:

I left Write Recovery Time at 12T in the BIOS and I think that it should be right because EVEREST has it at 18T for both.

**neo_rtr** · 12-29-2010, 10:27 AM

Question about settings, FSB, NB on a Q9650 running 2x Hd4780x2
Does it need to be higher?
I am trying to get to a 4Ghz. some of the settings i tried so far dont even boot.

Best so far CPu 1.36 rest on Auto 9x425 !! Get stucj during Vista boot 9x 430 . 9x445. 9x450. 9x460
with various FSB, Nb settings.

**leo27** · 12-29-2010, 04:02 PM

I have a similar problem to you neo, at above 432 it gets hard for me to get it even to post..I think we're doing the same thing wrong

Its annoying because I just want to try and get this Q6600 to 3.5Ghz and then I'll be happy

EDIT:I just got it to POST at 438 its was FSB VTT and GTL's that were screwing it up go figure

Thread: X48 Rampage Formula Preview.

Thread Tools

Search Thread

Rate This Thread

Display

Bookmarks

Bookmarks

Posting Permissions