Post by vaz on Jun 10, 2007 17:41:48 GMT
What a journey i've had with this MDD.
I've spent days and days trying to find a stable over-clock speed on these cpu's.
Steve here from Aquamac helped me a great deal and i've learnt so much about the balance needed for this to work.
My MDD mac is super stable and doesn't get restarted in weeks and months sometimes. The system and software is all tuned and optimized to free up as much cpu as i can.
The bus is still 133Mhz, i figure if i can find a stable speed past 1.33Ghz then i'll try the bus mod to 167Mhz..or is this where my problem lies? Bus getting too flooded and supersaturated with anything higher than it is now?
I have 2gb of ram all pc2700 or higher. I have an exhaust fan beside the ram and cpu. All bays are full with 5 hard drives and 1 dvd burner. All pci slots are full too except the one above the ati 9800 to help dissipate heat out of the case with the pci plate open.
My current and stable voltage is 9x at dual 1.33Ghz.
I've tried every voltage from 9.5x to 12x for dual 1.4 and 1.47 configurations. Some were more stable than others, holding out encoding and gaming and 15+ apps open doing filesharing etc.
But each one of these jumper settings cause a panic sooner or later. The longest it stayed on was overnight but it froze when i touched it in the morning, yet still semi responsive.
I've eliminated a ram issue, experimenting with 1 stick at a time out of the 6 i have here. BUT it has crashed on me running Xbench memory tests and Rember more than 1 occasion. But i've ignored those crashes because like i said i can push this to the max for weeks straight without it coming down on me.
So that's when i installed that exhaust fan beside the ram and cpu.
Heat is most likely the culprit here not keeping anything stable past dual 1.33Ghz.
I still need to get ram sleeves in the near future.
This is the panic.log i get when it crashes:
Unresolved kernel trap(cpu 0): 0x300 - Data access DAR=0x000000000987C4FF PC=0x0000000048008490
Latest crash info for cpu 0:
Exception state (sv=0x2D39BA00)
PC=0x48008490; MSR=0x02003000; DAR=0x0987C4FF; DSISR=0x42000000; LR=0x000AC33C; R1=0x00040000; XCP=0x0000000C (0x300 - Data access)
Backtrace:
0x41A2001C
backtrace terminated - frame not mapped or invalid: 0x74090400
Proceeding back via exception chain:
Exception state (sv=0x2D39BA00)
previously dumped as "Latest" state. skipping...
AND
panic(cpu 1 caller 0x000A4CC8): simple lock (0x0038E024) deadlock detection, pc=0x0003D894
Latest stack backtrace for cpu 1:
Backtrace:
0x000952D8 0x000957F0 0x00026898 0x000A4CC8 0x0003D894 0x000B4254 0x000AD9D4 0x000ABF48
Proceeding back via exception chain:
Exception state (sv=0x2D17D500)
PC=0x00208378; MSR=0x0000F030; DAR=0xE07FD000; DSISR=0x42000000; LR=0x00208454; R1=0xBFFFE6F0; XCP=0x00000024 (0x900 - Decrementer)
Also some other weird ones like
panic(cpu 0 caller 0x0021C8F4): btree_journal_modify_block_end: about to write corrupt node!
I think it actually caused some errors on 4 out of my 5 drive which were easily fixed with Disk Utility and all is good again back in dual 1.33 mode.
Things were acting strange when overclocked..like background and foreground apps quitting unexpectidly, a simple job locking up like reading on safari, mail crashing, dock crashing. It seemed to crash when the cpu's were idle and not working hard at all. I can tell because it freezes with my cpu monitor open always. My HD's never get hotter than 46deg Celsius next to the cpu. I'm also using the fan kext hack at 11deg sooner.
I'm really frustrated and about to give up here unless someone can suggest something else i haven't tried.
Please e-mail
Thanks
VaZ
I've spent days and days trying to find a stable over-clock speed on these cpu's.
Steve here from Aquamac helped me a great deal and i've learnt so much about the balance needed for this to work.
My MDD mac is super stable and doesn't get restarted in weeks and months sometimes. The system and software is all tuned and optimized to free up as much cpu as i can.
The bus is still 133Mhz, i figure if i can find a stable speed past 1.33Ghz then i'll try the bus mod to 167Mhz..or is this where my problem lies? Bus getting too flooded and supersaturated with anything higher than it is now?
I have 2gb of ram all pc2700 or higher. I have an exhaust fan beside the ram and cpu. All bays are full with 5 hard drives and 1 dvd burner. All pci slots are full too except the one above the ati 9800 to help dissipate heat out of the case with the pci plate open.
My current and stable voltage is 9x at dual 1.33Ghz.
I've tried every voltage from 9.5x to 12x for dual 1.4 and 1.47 configurations. Some were more stable than others, holding out encoding and gaming and 15+ apps open doing filesharing etc.
But each one of these jumper settings cause a panic sooner or later. The longest it stayed on was overnight but it froze when i touched it in the morning, yet still semi responsive.
I've eliminated a ram issue, experimenting with 1 stick at a time out of the 6 i have here. BUT it has crashed on me running Xbench memory tests and Rember more than 1 occasion. But i've ignored those crashes because like i said i can push this to the max for weeks straight without it coming down on me.
So that's when i installed that exhaust fan beside the ram and cpu.
Heat is most likely the culprit here not keeping anything stable past dual 1.33Ghz.
I still need to get ram sleeves in the near future.
This is the panic.log i get when it crashes:
Unresolved kernel trap(cpu 0): 0x300 - Data access DAR=0x000000000987C4FF PC=0x0000000048008490
Latest crash info for cpu 0:
Exception state (sv=0x2D39BA00)
PC=0x48008490; MSR=0x02003000; DAR=0x0987C4FF; DSISR=0x42000000; LR=0x000AC33C; R1=0x00040000; XCP=0x0000000C (0x300 - Data access)
Backtrace:
0x41A2001C
backtrace terminated - frame not mapped or invalid: 0x74090400
Proceeding back via exception chain:
Exception state (sv=0x2D39BA00)
previously dumped as "Latest" state. skipping...
AND
panic(cpu 1 caller 0x000A4CC8): simple lock (0x0038E024) deadlock detection, pc=0x0003D894
Latest stack backtrace for cpu 1:
Backtrace:
0x000952D8 0x000957F0 0x00026898 0x000A4CC8 0x0003D894 0x000B4254 0x000AD9D4 0x000ABF48
Proceeding back via exception chain:
Exception state (sv=0x2D17D500)
PC=0x00208378; MSR=0x0000F030; DAR=0xE07FD000; DSISR=0x42000000; LR=0x00208454; R1=0xBFFFE6F0; XCP=0x00000024 (0x900 - Decrementer)
Also some other weird ones like
panic(cpu 0 caller 0x0021C8F4): btree_journal_modify_block_end: about to write corrupt node!
I think it actually caused some errors on 4 out of my 5 drive which were easily fixed with Disk Utility and all is good again back in dual 1.33 mode.
Things were acting strange when overclocked..like background and foreground apps quitting unexpectidly, a simple job locking up like reading on safari, mail crashing, dock crashing. It seemed to crash when the cpu's were idle and not working hard at all. I can tell because it freezes with my cpu monitor open always. My HD's never get hotter than 46deg Celsius next to the cpu. I'm also using the fan kext hack at 11deg sooner.
I'm really frustrated and about to give up here unless someone can suggest something else i haven't tried.
Please e-mail
Thanks
VaZ