Help the newbie with his build!!

[zlojack]

Hey Talonman, I've been admiring your thread here plus the Maximus Formula WC thread (I am VERY tempted to go to the WC side of the force)

Anyway, my experience with the Q6600 with the Maximus Formula so far is that I could get a higher FSB and more stable RAM with the 400 strap than the 333. For some reason either my RAM or my mobo didn't like the 333.

Even now when I'm testing my E8400, the RAM doesn't work on anything but AUTO.
My Q6600 was at 3.6 on air, but it would max out around 65-70 degrees at load. 40C must be nice!

[Talonman]

Thanks zlojack.

I am just too happy with the way my temps are looking.

I think water cooling rules, and it is allot of fun too. It is not the cheepest option, but I think that they say water transferes heat 100 X better than air I believe... It's just gotta help. I'm lovein 40C at my CPU's worst. It is giving me some peace of mind while I am tuning er up, and larger room for error...

I will be switching from 333 to 400 to test that out too! thanks.

[zlojack]

Well, I had success with the 400 strap so I hope it works for you.

[Talonman]

It looks like the 400 Strap will do well for me. I just updated the BIOS and so far OCCT, has no problem with it. Updating my settings!

If 400 continues to be stable, I will try to eak my FSB back up toward 425 again...
Update: North Bridge Strap of 400, with my 415 FSB ran clean on OCCT for 1 hour.
Keeping my new North Bridge Strap setting of 400...
Tried 420 FSB, and OCCT ran longer than it did with the 333 NB Strap, but still received an error code 1
Tried 418 FSB, ran for 20 minutes, but also received error code 1.
Tried 416 FSB, ran 1 hour clean OCCT run. Tried it in Prime95, and one core failed in about 20 minutes.
Set FSB back to 415, and fired Prime95 up. 415 had not been tested yet using Prime95.
Prime95 ran for 45 minutes error free, but then had 1 failure.
Set FSB back to 414. This still gives me 3.726GHz on the CPU.
Tested FSB 414, and a North Bridge Strap of 400 on Prime95.
It successfully completed 1.5 hours of Prime95 error free....

It also drove my CPU's temp's to a new high! 3.726GHz
This shot was taken right before I manually shut down Prime95, after successfully completing 1.5 hours of hard labor.


It appears my Maximus won't give me much more than 3.726GHz, with a 414 FSB, on a NB Strap 333 or 400. I think I favor the 400 and am keeping that so far.

I can't complain. My system is now running at 3.7GHz, with a 414 FSB speed, and max CPU temp around 48C!

These are my current best BIOS 907 settings: Always looking for advice.
Under Extreme Tweaker
Ai Overclock Tuner : Manual
OC From CPU Level Up : Auto
CPU Ratio Setting : 9
FSB Frequency : 414
FSB Strap to North Bridge : 400
PCI-E Frequency: 110
DRAM Frequency: DDR2-828MHz
DRAM Command Rate : 2T
DRAM Timing Control: Manual
CAS# Latency : 5
RAS# to CAS# Delay : 5
RAS# Precharge : 5
RAS# ActivateTime : 18
RAS# to RAS# Delay : Auto
Row Refresh Cycle Time : Auto
Write Recovery Time : Auto
Read to Precharge Time : Auto

Read to Write Delay (S/D) : Auto
Write to Read Delay (S) : Auto
Write to Read Delay (D) : Auto
Read to Read Delay (S) : Auto
Read to Read Delay (D) : Auto
Write to Write Delay (S) : Auto
Write to Write Delay (D) : Auto
DRAM Static Read Control: Disabled
Ai Clock Twister : Auto
Transaction Booster : Auto

CPU Voltage : 1.475 (BIOS)
CPU PLL Voltage : 1.50 (BIOS)
North Bridge Voltage : 1.57 (BIOS)
DRAM Voltage : 1.80(BIOS)
FSB Termination Voltage : 1.44 (BIOS)
South Bridge Voltage : 1.05 (BIOS)
Loadline Calibration : Enabled
CPU GTL Reference : 0.63x
North Bridge GTL Reference : 0.67x
DDR2 Channel A REF Voltage : Auto
DDR2 Channel B REF Voltage : Auto
DDR2 Controller REF Voltage : Auto
SB 1.5V Voltage : 1.50 (BIOS)

NB LED Selection : NB Volt
SB LED Selection : SB Volt
CPU LED Selection : CPU Volt
Voltiminder LED : Enabled

CPU Spread Spectrum : Disabled
PCIE Spread Spectrum : Disabled

Under Advanced
And CPU Configuration:
CPU Ratio Setting : 9
C1E Suppport : Disabled
CPU TM Function : Disabled
Vanderpool Technology : Disabled
Execute Disable Bit : Disabled
Max CPUID Value Limit : Disabled

Under Advanced
And USB Configuration :
USB Functions: Enabled
USB 2.0 Controller : Enabled
BIOS EHCI Hand-off : Enabled
Legacy USB Support : Disabled
Port 64/60 Emulation : Disabled

FYI - I just spotted this post:

What is the rated voltage of your ram and what do you have dram voltage set to?

PCI-E Frequency: 100
Ai Clock Twister : Light
Transaction Booster : Enabled (FSB Strap to North Bridge = 400)
BOOST: 0

OR
Transaction Booster : Disabled (FSB Strap to North Bridge = 333)
BOOST: 0

DDR2 Controller REF Voltage : DDR2
CPU Spread Spectrum : Disabled
PCIE Spread Spectrum : Disabled

This tid bit of information may or may not help you.

Due to me now using FSB Strap to North Bridge of 400, I think I am going to try the PCI-E Frequency: 100
Ai Clock Twister : Light,Transaction Booster : Enabled, and BOOST: 0.

My system would not pass the RAM check with Transaction Booster enabled.

Testing AI clock twister set to : Light
DDR2 Controller set back to DDR2_REF
Took CPU Voltage back to 1.45v
Took NB voltage down to 1.49v in BIOS.

[Talonman]

Doing some fine tuning...

I walked the NB voltage down to 1.49v before starting to get system errors. I then moved the voltage back up to 1.51v and successfully primed for 1.5 hours. 1.51v will be kept for my new NB voltage.

AI Clock Twister changed from Auto to Light worked fine. More info is needed on the subjuct, but for now I am keeping the Light setting.

DDR2 Controller set from Auto to DDR2_REF worked fine, and it was posted in the Maximus Part 2 thread, that having it set to DDR2_REF might add to system stability? I am keeping the DDR2_REF setting.

Walking down the CPU voltage it takes for me to successfully run 3.7GHz. CPU voltage was set at 1.57v in the BIOS. 1.46250v has already passed a 1.5 hour Prime95 test error free. I am still continuing the process until Prime95 received an error at 1.45v's. I set my CPU voltage back up to 1.46250v and am using it for my keeper setting.

I posted my OC'ing info in the Maximus Part II Thread, but wanted to include it in my build thread too.
This is what I tried to get from 3.7GHz to 3.8GHz:

Test - FSB FREQ From 414 to 416 using Prime95 - Result: Core 2 Hardware Failure in 4 minutes.
(Verified current setting still cause errors above a 414 FSB Freq.)

Test - Kept 416 FSB FREQ, but increased FSB Term volts from 1.44 to 1.46 - Result: Core 3 Hardware Failure in 4 minutes.

Test - Kept 416 FSB and FSB Term 1.46v, but also changed:
CPU Voltage : From 1.46250v to 1.506.
CPU PLL Voltage : From 1.50v to 1.56.
Result: Prime95 ran error free for 10 minutes.

Test - Restored FSB Term volts to 1.44
Kept - CPU Voltage of 1.506, and CPU PLL Voltage of 1.56, but also increased FSB FREQ to 418.
Result: Prime95 ran error free for 10 minutes.
(I do not believe upping the Term volts was helping me with my FSB wall. I returned them to what the were and still ran clean with the upped CPU voltages.)

Test - Kept CPU Voltage of 1.506, and CPU PLL Voltage of 1.56, but also increased FSB FREQ to 420.
Result: Prime95 ran error free for 10 minutes.
(Short test but looking good)

Test - Kept CPU Voltage of 1.506, and CPU PLL Voltage of 1.56, but also increased FSB FREQ to 430.
Result: System booted into Vista fine, and had the CPU clocked to 3.87GHz, but Core 3 had a Hardware Failure in 3 minutes.

Test - Kept CPU Voltage of 1.506, and CPU PLL Voltage of 1.56, but also lowered FSB FREQ to 425.
Result: Prime95 ran error free for 10 minutes!

I wanted to see if I could successfully run a 430 FSB, but only up the CPU PLL Voltage, not the CPU voltage:
Test - Kept CPU Voltage of 1.506, but upped CPU PLL Voltage from 1.56 to 1.62, and returned the FSB FREQ to 430.
Result: Hardware Failure in 3 minutes.

Tried again by upping the CPU PLL Voltage more!
Test - Kept CPU Voltage of 1.506, but upped CPU PLL Voltage from 1.62 to 1.64, and kept the FSB FREQ to 430.
Result: Hardware Failure in 3 minutes.

Action Taken: Returned the CPU PLL Voltage to 1.56v

Conclusion: My FSB Frequency wall of 414 was being directly caused by the voltage supplied to the CPU's core. My CPU Voltage setting of 1.46250v (BIOS), only will get me my 3.7GHz on my Q6600.

My new CPU Voltage setting of 1.46250v (BIOS), looks to be good for about 3.8GHz. Everest does read lower than the BIOS'es setting, but how high can I feel good about setting my voltage to in the BIOS for my Q6600?

I think I could have 4.0GHz if I am willing to overvolt my CPU a bit more?

[dude_500]

My 3.6ghz load temp is around 55*

Is my apogee GT just out dated now or something? I have BIP III and MCR220 so it should be enough radiator power

[Talonman]

I can't say for sure?

Too new into the game myself!

Note to get 3.8GHz: I had to bump up my CPU core voltage to 1.506v in the BIOS, but Everest is only reporting 1.49 volts, so I may not be actually over-volting my CPU as much as I may think...

Keeping my new CPU voltage of 1.506v in the BIOS, I tried to run Prime95 using a FSB Frequency of:
428 - Result: Prime error in 4 minutes.
427 - Result: Prime error in 4 minutes.
426 - Result: Prime error in 10 minutes.
425 - Result: Prime error in 23 minutes!
424 - Result: Prime stable for 1 Glorious hour!! (Keeping 424 for now)

This makes my memory run at 848MHz, and gives me a Core speed of 3816.3MHz.
Not by much, but still enouigh to officially get me into the 3.8GHz club!
Prime doing it's thing crossing the hour mark error free. Still looking too.


These are my current best BIOS 907 settings for 3.81GHz: Always looking for advice.

Under Extreme Tweaker
Ai Overclock Tuner : Manual
OC From CPU Level Up : Auto
CPU Ratio Setting : 9
FSB Frequency : 424
FSB Strap to North Bridge : 400
PCI-E Frequency: 110
DRAM Frequency: DDR2-848MHz
DRAM Command Rate : 2T
DRAM Timing Control: Manual
CAS# Latency : 5
RAS# to CAS# Delay : 5
RAS# Precharge : 5
RAS# ActivateTime : 18
RAS# to RAS# Delay : Auto
Row Refresh Cycle Time : Auto
Write Recovery Time : Auto
Read to Precharge Time : Auto

Read to Write Delay (S/D) : Auto
Write to Read Delay (S) : Auto
Write to Read Delay (D) : Auto
Read to Read Delay (S) : Auto
Read to Read Delay (D) : Auto
Write to Write Delay (S) : Auto
Write to Write Delay (D) : Auto
DRAM Static Read Control: Enabled
Ai Clock Twister : Strong
Transaction Booster : Auto

CPU Voltage : 1.506v (BIOS)
CPU PLL Voltage : 1.56v (BIOS)
North Bridge Voltage : 1.51v (BIOS)
DRAM Voltage : 1.80v (BIOS)
FSB Termination Voltage : 1.44v (BIOS)
South Bridge Voltage : 1.05v (BIOS)
Loadline Calibration : Enabled
CPU GTL Reference : 0.63x
North Bridge GTL Reference : 0.67x
DDR2 Channel A REF Voltage : Auto
DDR2 Channel B REF Voltage : Auto
DDR2 Controller REF Voltage : DDR2_REF
SB 1.5V Voltage : 1.50v (BIOS)

NB LED Selection : NB Volt
SB LED Selection : SB Volt
CPU LED Selection : CPU Volt
Voltiminder LED : Enabled

CPU Spread Spectrum : Disabled
PCIE Spread Spectrum : Disabled

Under Advanced
And CPU Configuration:
CPU Ratio Setting : 9
C1E Suppport : Disabled
CPU TM Function : Enabled
Vanderpool Technology : Disabled
Execute Disable Bit : Disabled
Max CPUID Value Limit : Disabled

Under Advanced
And USB Configuration :
USB Functions: Enabled
USB 2.0 Controller : Enabled
BIOS EHCI Hand-off : Enabled
Port 64/60 Emulation : Disabled
Legacy USB Support : Enabled

I am not sure how much voltage I want to give my CPU 24 X 7. For now I think I will keep my 3.8GHz, and wait for my GPU...

[Buckeye]

So you don't go over 40c-43c at 1.5 vcore and 3.8ghz running with the Keg while prime95 is running ?

[Talonman]

Correct!!

Even after a full hour of hard labor.

I also now believe that Prime95 runs your temps up higher than OCCT, after 1 full hour of processing time too.
I think I used the max heat generating program for the above test...

[Buckeye]

HAHA That is Great !

Seems like the Keg works really sweet

I know at 3.8ghz and 1.5 vcore my Freezone was getting into the 60c's so I backed it off.

[Talonman]

Between the WaterKegIII Xtreme, water block selection, D-TEK FuZion Nozzle Kit, D-TEK FuZion Pro Mount, Lapping the CPU, and MX-2 used for my TIM, something paid off!!

Even my SB sink/fan combo seems to do well.

My 62F basement also is helping I'm sure...

[Buckeye]

HAHA yeah it did. Gratz Talon !!

[Talonman]

Thanks man!

I have the proper system now, to keep up with a speedy GPU!

I am going to tame Crysis some day, with max everything, at 1920 x 1200!

Note my signature update...
Processor: Q6600 G0 @ 3.8GHz

It just needed to be done. You can't have any street cred on XtremeSystems running stock!

[Talonman]

Just for the record:

One thing I noted about Prime95: If you look at my 3.7GHz screen shot I hit 48C. Prime95 during the 1HR to 1.5 hour timeframe, does some extra heat generating calculations or something....
During my 3.7GHz test I went 1.5 hours of Prime95 when I took that screen shot. It drove my temps to a new high, and mostly the last few C were in the last 1/2 hour.

As I recall my high temps were around 40C at the hour mark in that test too.

You can drive your temps up a few C more if you go to 1.5 hours I think, over just running 1.

Not really sure.

Some of the boys in the IP35-Pro thread were discussing about how letting Prime95 run for 24 hours or so, may actually dammage your CPU due to excessive stress. Some further think that no real life situaction, even demanding games, will ever stress your CPU like hours of Priming does.

I tend not to like to Prime for hours and hours, and buy into the thinking that 1 solid hour of Prime95 is stable enouigh for a gameing rig like mine...

I also am of the mindset that even Crysis won't drive my temps up any higher than 1 full hour running Prime95. It is a brutal program! To me the numbers still stand as a valid measuring stick to what I can expect my 'real world' max temp to be, playing any game.

I think given extended time, Prime95 could take them a wee bit higher, but also feel only Prime would do so. Not a huge concern to me...
I just gotta believe my cool running 3.8GHz Quad system, is just waiting for a Bad Boy GPU to eat up some searious games now.
I have my 1000W PSU just waiting to feed something nasty too.

Fact: The Nvidia 256MB 8600GT is currently the fastest GPU I have ever owned...
I am going to be in for a ride!! Where is that 9800GTX anyway?

[Buckeye]

I ran BFG 8800 GTX OC's in SLI on my striker for awhile. Maybe it was then drivers where not that great but I was not to impressed with SLI then, now just running one 8800 GTX.

But that said...

With what ever Nvidia releases I am really looking at a single card solution that can go TRI SLI. So if they do actually make a 9800 GTX or something even better that is what I am planning on setting up. I am not that crazy about that other new card 9800GX2 or what ever it is as I prefer a single GPU card.

[Talonman]

Me too..

[Buckeye]

So if you put your PC into a an aquarium and filled it with mineral oil and used the Keg to circulate the oil you might have something like this. Crazy !! Forget blocks HAHA

http://www.techeblog.com/index.php/t...-system-video-

[disruptfam]

any luck on that 4ghz mate?

[Talonman]

OK.. I loaded my saved 3.9GHz BIOS profile back in, and did a 20 minute barn burner run!!
3.9GHz took 1.55 volts in the BIOS, 4.0 might want 1.6v?

Recap of my findings:
So far With the Maximus and my Q6600 GO:
My CPU Voltage setting of 1.46250v (BIOS), will allow error free Priming with a FSB Frequency 414 giving me 3.7GHz with my 400 NB Strap.
My CPU Voltage setting of 1.506v (BIOS), will allow error free Priming with a FSB Frequency 424 giving me 3.81GHz with my 400 NB Strap.
My CPU Voltage setting of 1.51875v (BIOS), will allow error free Priming with a FSB Frequency 428 giving me 3.85GHz with my 400 NB Strap.
My CPU Voltage setting of 1.51875v (BIOS) starting taking errors 4 minutes into prime with a FSB Frequency 432 giving me 3.8888GHz with my 400 NB Strap.

I did successfully prime 3.9GHz with a FSB Frequency 435 for 45 minutes, but stopped the test.
For my system it requires me to set in the BIOS a CPU core voltage of 1.55000v's, and Everest reads 1.53v.
You also must bump your CPU PLL to 1.60v's.
1.55 Core and only 1.56 PLL would not Prime 3.9GHz error free.

20 minutes of Prime95 at 3.9GHz...

[Talonman]

So if you put your PC into a an aquarium and filled it with mineral oil and used the Keg to circulate the oil you might have something like this. Crazy !! Forget blocks HAHA

http://www.techeblog.com/index.php/t...-system-video-

That is pritty wild!

Kind a messy though on adding things to the MOBO....

[Talonman]

More fine tuning was done on my every day 3.8GHz Bios settings.

I primed for 2 hours using my above 3.8GHz settings, and encountered an error on 1 core. My guess was it probably was just Core voltage issue, but remember when Priming 3.9GHz I needed a wee bit more CPU PLL voltage too.

Allot of the boy's are running more NB voltage than me and my 1.51v's, so I took that up a notch too to 1.55V's.

I also moved my South bridge voltage from 1.05v's to 1.075v's for no good reason.

I cutt Prime95 loose again to see if my 3.8GHz would still have an error in 2 hours. My result was an error free 4 hour run, and me adopting these new settings.
This is what I currently run 24-7 now...

My 4hour error free run using my 3.8GHz settings. Room temp was a brisk 55°F:


These are my current best BIOS 907 settings for 3.81GHz:

Under Extreme Tweaker
Ai Overclock Tuner : Manual
OC From CPU Level Up : Auto
CPU Ratio Setting : 9
FSB Frequency : 424
FSB Strap to North Bridge : 400
PCI-E Frequency: 110
DRAM Frequency: DDR2-848MHz
DRAM Command Rate : 2T
DRAM Timing Control: Manual
CAS# Latency : 5
RAS# to CAS# Delay : 5
RAS# Precharge : 5
RAS# ActivateTime : 18
RAS# to RAS# Delay : Auto
Row Refresh Cycle Time : Auto
Write Recovery Time : Auto
Read to Precharge Time : Auto

Read to Write Delay (S/D) : Auto
Write to Read Delay (S) : Auto
Write to Read Delay (D) : Auto
Read to Read Delay (S) : Auto
Read to Read Delay (D) : Auto
Write to Write Delay (S) : Auto
Write to Write Delay (D) : Auto
DRAM Static Read Control: Enabled
Ai Clock Twister : Strong
Transaction Booster : Auto

CPU Voltage : 1.51250v (BIOS)
CPU PLL Voltage : 1.58v (BIOS)
North Bridge Voltage : 1.55v (BIOS)
DRAM Voltage : 1.80v (BIOS)
FSB Termination Voltage : 1.44v (BIOS)
South Bridge Voltage : 1.075v (BIOS)
Loadline Calibration : Enabled
CPU GTL Reference : 0.63x
North Bridge GTL Reference : 0.67x
DDR2 Channel A REF Voltage : Auto
DDR2 Channel B REF Voltage : Auto
DDR2 Controller REF Voltage : DDR2_REF
SB 1.5V Voltage : 1.50v (BIOS)

NB LED Selection : NB Volt
SB LED Selection : SB Volt
CPU LED Selection : CPU Volt
Voltiminder LED : Enabled

CPU Spread Spectrum : Disabled
PCIE Spread Spectrum : Disabled

Under Advanced
And CPU Configuration:
CPU Ratio Setting : 9
C1E Suppport : Disabled
CPU TM Function : Enabled
Vanderpool Technology : Disabled
Execute Disable Bit : Disabled
Max CPUID Value Limit : Disabled

Under Advanced
And USB Configuration :
USB Functions: Enabled
USB 2.0 Controller : Enabled
USB 2.0 Controller Mode: HiSpeed
BIOS EHCI Hand-off : Enabled
Port 64/60 Emulation : Disabled
Legacy USB Support : Enabled

[Buckeye]

Nice !!

Getting close to 4.0

[zlojack]

Hey Talonman... I'm looking at that SB cooler you glued on there.

Is there really no way to put it on without the adhesive? I'm somewhat leery about attaching it permanently...

[Talonman]

I really can't speak to that directly, as I think the sinks low profile won't interfear with a video card if you opt to run 2. The fan would have to go for sure in a dual GPU config. I don't hold much respect for the stock SB sink. We have had a few reports of guys feeling pritty darn strongly that the NB was raising the SB temp via the connecting heat pipe. Our NB is a flame thrower, and is not any friend of the SB aiding in cooler temps. I think it hurts them in actuality.

I don't believe a dedicated fan is required to have low SB temps with that heat sink. You might be able to run a bolt out both existing holes in the MOBO, and run a clamp accrost the top of the sink mabey?? Not sure.

I do grove on the idea of the bond between Chip and Sink I now have. Using the flashlight for some extra weight while the bond set-up worked well.
Performance first I say.

[zlojack]

I really can't speak to that directly, as I think the sinks low profile won't interfear with a video card if you opt to run 2. The fan would have to go for sure in a dual GPU config. I don't hold much respect for the stock SB sink. We have had a few reports of guys feeling pritty darn strongly that the NB was raising the SB temp via the connecting heat pipe. Our NB is a flame thrower, and is not any friend of the SB aiding in cooler temps. I think it hurts them in actuality.

I don't believe a dedicated fan is required to have low SB temps with that heat sink. You might be able to run a bolt out both existing holes in the MOBO, and run a clamp accrost the top of the sink mabey?? Not sure.

I do grove on the idea of the bond between Chip and Sink I now have. Using the flashlight for some extra weight while the bond set-up worked well.
Performance first I say.

I hear you!

Something to consider...