E6700ES @4780MHz under 2-Stager

[before]

E6700ES @4951MHz under 2-Stager

SYSTEM
CPU: Quent's E6700ES 5B1
MoBo: Asus P5B-DLX WiFi
RAM: Super Talent T800UX1GC4
OS: Windows XP Professional SP2
HDD: Western Digital IDE 80Gb

UPDATE

5182MHz

http://valid.x86-secret.com/show_oc?id=134250

4951MHz

Vcore= 1.81V (higher volts didn't give any improvement)
Vmch= 1.71V (it couldn't go over without making reboots)
Vddr= 2.70V (didn't try with more volts )

Normal mode with Green Olive scheme (Reg keys* + Maxmem=600): best run was achieved in 10.390s

*Among others, I've used these dword values:

[HKEY_LOCAL_MACHINE\System\CurrentControlSet\Control\Session Manager\Executive]
"AdditionalCriticalWorkerThreads"=dword:00000000
"AdditionalDelayedWorkerThreads"=dword:00000000
"DefaultNumberOfWorkerThreads"=dword:00000001

[HKEY_LOCAL_MACHINE\System\CurrentControlSet\Control]
"PriorityControl"=dword:00000002 (Foreground windows have quantums three times as long as other windows)

-----------------------------------------------------
-----------------------------------------------------

Cooling: H2O

I've spent the last two days trying to select the best tweaks for Super PI. In fact, benching Super PI with C2D is very different than with A64. I was quite confident with my good old tweaks but I must confess that I'm a noob now for tweaking Super PI while running a C2D.
The CPU was cooled by an homemade watercooling system and only needed 1.4Vcore to operate within @3600MHz to 3680MHz.

Normal mode with Olive Green color scheme (Tweaks: modified registery keys + Maxmem=600): best run achieved in 13.797s

Normal mode (Tweaks: modified registery keys + No page file + Maxmem=600 + Copywaza=1.6Gb): single run achieved in 14m 37.844s

-----------------------------------------------------
-----------------------------------------------------

Cooling: Cascade (R290+R170 refrigerants, 8082+7950Btu/h rotaries compressors, @ about -100°C)

But today I've put the bad baby under 2-Stager to begin serious things... My motherboard is fully volt-modded but I didn't tune the Vmod trimmers yet and only set voltage settings to the max instead of the CPU which vcore has been set to 1.60V 'only'.

Take a look a the temp

Normal mode with Olive Green color scheme (Tweaks: modified registery keys + Maxmem=600): best run achieved in 10.703s

http://valid.x86-secret.com/show_oc?id=132820

I've to find more useful tweaks but the next step is to overclock a little bit more...

 

[greenmaji]

Great work so far before

Now, more Vcore, Vdimm and tweaks
Unless you have a memory maximizer on hold from hipro-tech Then you could ignore the Vdimm coment

 

[dominick32]

Wow dude!
Awesome work. 2-Stage cascade with better than DICE temperatures. Keep working at 5 ghz. I am curious, what do your loaded temps look like at 4.7??

 

[greenmaji]

2-Stage cascade with better than DICE temperatures.

before didnt play around Ethane and big old rotarys = cascade
Not quite but boy does it perform

I've never asked but I hope you have a blow-off valve on the Ethane/second stage

 

[dominick32]

If I remember correctly. These were Epox4life DICE temps in the bios:

So, this guys 2 stage cascade performs better than some DICE setups. That thing is killer!!!

 

[greenmaji]

Ethane's boiling point it FREAKING LOW!! My chart says -88.9C, his rotaries must be running the system under a vacume HAHA
But its a HIGH PREASURE GAS!! Some builders won't even use it.

The cascade Jinu built for sale IIRC he used R23 in the second stage for safty reasons rather then performance.

 

[Gautam]

Seems like the two stager is performing a bit better than DI considering the vcore and idle temps.

Your 1M time seems fine, but the 32M at 3.6GHz is off...this is just 400x9 1:1. This might be whatever problem Ross was having with his Pi times on the P5B. Seems like its a step 5 ES thing. Try booting at 400MHz and use clockgen to increase.

 

[before]

Thank you so much guys

loading temp with the chip operating at [email protected] was -80°C... only 9°C from the idle one. In order to have -72.5°C reading into the bios I've to push the chip at [email protected] Just a guess because if I correctly remember, at [email protected] the bios temp is about -75°C. Windows reading is about 3°C to 4°C lower. At stock, the bios temp is -92°C to -93°C but -96°C with Speedfan.

I've got something weird while pushing up the CPU over about 4.8GHz with clockgen (Spi32M is ok at 4.7GHz; Spi1M can also pass at 4.8GHz). I don't know what the matter is about? It might be the RAM (but CL5 and 2.84V -bios Vddr estimate- didn't help either) or the chipset because I can't believe the chip is failing even with 1.70Vcore as set in the bios.

I'm not sure about the voltages given by my DMM so I'll buy a better one in order to safely tune the trimmers.

@Gautam, I'll try what you've suggested Thx bro

Here are some of the registery keys I'm used to play with:

[HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows\Curr entVersion\Explorer]
“AlwaysUnloadDll”=“1”

[HKEY_CURRENT_USER\Control Panel\Desktop]
“MenuShowDelay”=“0”
“FontSmoothing”=“2”
“AutoEndTasks”=“1”
“HungAppTimeout”=“40”
“WaitToKillAppTimeout”=“200”

[HKEY_USERS\.DEFAULT\Control Panel\Desktop]
“MenuShowDelay”=“0”
“FontSmoothing”=“2”
“AutoEndTasks”=“1”
“HungAppTimeout”=“40”
“WaitToKillAppTimeout”=“200”

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Session Manager\Memory Management]
"ClearPageFileAtShutdown"=dword:00000001
"DisablePagingExecutive"=dword:00000001
"IoPageLockLimit"=dword:00010000
"LargeSystemCache"=dword:00000001
"SystemPages"=dword:00183000
“PhysicalAddressExtension”=dword:00000000
"SecondLevelDataCache"=dword:00001000 (only used if the system fails to identify CPU actual cache)

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Contro l\Session Manager\Memory Management\PrefetchParameters]
“VideoInitTime”=dword:00000290 or something else (looks like useless; it changes after each boot, that’s puzzeling me!!)
“EnablePrefetcher”=dword:00000001
“EnableSuperfetch”=dword:00000001

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Contro l\FileSystem]
"ContigFileAllocSize"=dword:00000000 (looks like useless after Win 95)
"NtfsDisableLastAccessUpdate"=dword:00000000
"NtfsDisable8dot3NameCreation"=dword:00000000
"Win31FileSystem":dword:00000000
"Win95TruncatedExtensions":dword:00000001

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Contro l\PriorityControl]
"Win32PrioritySeparation"=dword:00000000
"IRQ8Priority"=dword:00000000

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Contro l\Session Manager\Executive]
"AdditionalCriticalWorkerThreads"=dword:00000000 (The less is the best)
"AdditionalDelayedWorkerThreads"=dword:00000000 (The less is the best)
"DefaultNumberOfWorkerThreads"=dword:00000000 (The less is the best)

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\PriorityControl]
"PriorityControl":dword:00000002

The last 2 ones are used in order to have a Windows 2k3 server start up instead of the classical Win XP start up. This tweak was very nice and really time saving with A64. It might be useless now... I don't know.

[HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows NT\CurrentVersion\Winlogon]
"LogonType"=dword:00000000

[HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows NT\CurrentVersion\Winlogon\SpecialAccounts\UserLis t]
"before"=dword:00000000

 

[{FKR}Loki]

Very nice Temps and OC.

 

[before]

I can reach 5GHz (and 5.2GHz too ) but can't pass a single Spi1M even with more than 1.81Vcore

Vcore= 1.81V (higher volts didn't give any improvement)
Vmch= 1.71V (it couldn't go over without making reboots)
Vddr= 2.70V (didn't try with more volts )

Normal mode with Green Olive scheme (Reg keys* + Maxmem=600): best run was achieved in 10.390s

*Among others, I've used these dword values:

[HKEY_LOCAL_MACHINE\System\CurrentControlSet\Control\Session Manager\Executive]
"AdditionalCriticalWorkerThreads"=dword:00000000
"AdditionalDelayedWorkerThreads"=dword:00000000
"DefaultNumberOfWorkerThreads"=dword:00000001

[HKEY_LOCAL_MACHINE\System\CurrentControlSet\Control]
"PriorityControl"=dword:00000002 (Foreground windows have quantums three times as long as other windows)

 

[before]

I hope I'll be able to do more... BTW I've tried this morning and after starting, the system froze into the bios and refused to reboot again.

Then I've turned off the cascade and unmounted the evaporator and insulation to find ice and water around the socket. I hope the motherboard and CPU (which is not mine) are still OK.

 

[dominick32]

Well sir. I suggest you get a 5 ghz validation and go straight to here:
http://www.ocforums.com/showthread.php?t=480042&highlight=5ghz

Do not pass go, do not collect $200 dollars.

 

[before]

I already had one near 5.2Ghz... I'll go there after the next benching

PS: you'r lucky to live in NY... I really love this city

 

[Gautam]

Quent's CPU seems amazing. Paired with your 2 stager, its quite a dangerous combo.

 

[before]

Yeah right

Quent has just told me he'll try to find a X6800 for my cascade This guy is an angel! He wished to send me a bottle LN2 too, but it seems quite difficult to post such an item by the postal service.

So, I've tried yesterday to bench @5Ghz and improved the suicide screen as well. BTW my Windows XP looked like corrupted and it was too late in the night to continue.

Here's the actual max screen taken by CPU-Z. I've also a validation @5.2GHz but it is corrupted and can't be uploaded to the CPU-Z database.

http://valid.x86-secret.com/show_oc?id=134250

 

[HungryForHertz]

God this creates a party in my pants. Nice work.

 

[before]

hehe Hungry

Well, I've performed some Spi1M action today with 9x instead of 10x. It scaled nicely.

--- [10.985s] -----> 4644MHz @1.67V / 516MHz 4-4-3-4 @2.55V / Vmch = 1.61V

http://valid.x86-secret.com/show_oc?id=134568

--- [10.890s] -----> 4683MHz @1.67V / 520MHz 4-4-3-4 @2.55V / Vmch = 1.61V

http://valid.x86-secret.com/show_oc?id=134572

--- [10.782s] -----> 4732MHz @1.67V / 526MHz 4-4-3-4 @2.55V / Vmch = 1.61V

http://valid.x86-secret.com/show_oc?id=134575

--- [10.687s] -----> 4781MHz @1.70V / 531MHz 4-4-3-4 @2.61V / Vmch = 1.64V

http://valid.x86-secret.com/show_oc?id=134578

--- [10.594s] -----> 4821MHz @1.73V / 536MHz 4-4-3-4 @2.61V / Vmch = 1.64V

http://valid.x86-secret.com/show_oc?id=134580