It is always dangerous to say you are experienced or expert -- especially if you are having a problem you can't resolve for yourself -- but I have built many overclocked systems and am certainly experienced if not expert. I have a well proven protocol to creep up on the best stable configuration. Not mine I must admit, but used by me for years, involving Memtest and Prime95.
The system in question has two sticks of Corsair XMS 3200 400 MHz CL2 platinum series memory which will run Memtest 86 for 24 hours with no errors at 3.2 GHz at 2.8v and timings of 2.5,3,3,7. I have run it overnight at those settings several times in the course of troubleshooting the problem.
The CPU is a Prescott 2.4 GHz which overclocks to 3.2 GHz with a core of 1.55v and runs at 32 degrees under no load. It is well cooled. The OS is Windows XP with SP2 installed.
I have run Prime95 for a few hours with no problem (except as noted later) and also computed pi to several million places running superpi 1.5
The problem is that the system reboots on it's own, even if no programs are running. Just sitting there with the desktop showing, it will randomly reboot. Since it isn't the memory and the CPU is not running hot, the natural assumption is that the system is overclocked beyond what this particular CPU is good for. OK, so I backed off the FSB in 10 MHz steps all the way back down to the default 200 MHz. The random reboots still occur. Since they are random I can't say if the frequency with which they occur is less or not, but they still occur. Since power supplies are a natural source of problems, I swapped in a new Thermatake 480w supply. Still the same problem.
I unchecked the restart on failure option so I could look at the event log. The only red checked items are for dl1 not installed on startup. This is associated with Zone Alarm so I contacted their tech support and was told this was normal (?). At any rate they did not think it was associated in any way with the random reboots.
The problem is not due to memory I am pretty sure since 24 hours of Memtest (at the highest FSB setting of 266 MHz) have failed to cause a single error. But I should mention that I have loosened the memory timings way up, just to troubleshoot, with no effect. It is certainly not due to heat since even the Northbridge is running at only warm to the touch and the CPU is as cool as a cucumber. As mentioned above, the reboots occur with no overclocking at all. So does anyone have any idea of what could be the problem?
It is a good system -- when it runs -- computing one million digits of pi in 42 seconds.
The system in question has two sticks of Corsair XMS 3200 400 MHz CL2 platinum series memory which will run Memtest 86 for 24 hours with no errors at 3.2 GHz at 2.8v and timings of 2.5,3,3,7. I have run it overnight at those settings several times in the course of troubleshooting the problem.
The CPU is a Prescott 2.4 GHz which overclocks to 3.2 GHz with a core of 1.55v and runs at 32 degrees under no load. It is well cooled. The OS is Windows XP with SP2 installed.
I have run Prime95 for a few hours with no problem (except as noted later) and also computed pi to several million places running superpi 1.5
The problem is that the system reboots on it's own, even if no programs are running. Just sitting there with the desktop showing, it will randomly reboot. Since it isn't the memory and the CPU is not running hot, the natural assumption is that the system is overclocked beyond what this particular CPU is good for. OK, so I backed off the FSB in 10 MHz steps all the way back down to the default 200 MHz. The random reboots still occur. Since they are random I can't say if the frequency with which they occur is less or not, but they still occur. Since power supplies are a natural source of problems, I swapped in a new Thermatake 480w supply. Still the same problem.
I unchecked the restart on failure option so I could look at the event log. The only red checked items are for dl1 not installed on startup. This is associated with Zone Alarm so I contacted their tech support and was told this was normal (?). At any rate they did not think it was associated in any way with the random reboots.
The problem is not due to memory I am pretty sure since 24 hours of Memtest (at the highest FSB setting of 266 MHz) have failed to cause a single error. But I should mention that I have loosened the memory timings way up, just to troubleshoot, with no effect. It is certainly not due to heat since even the Northbridge is running at only warm to the touch and the CPU is as cool as a cucumber. As mentioned above, the reboots occur with no overclocking at all. So does anyone have any idea of what could be the problem?
It is a good system -- when it runs -- computing one million digits of pi in 42 seconds.
Facebook
Twitter