- Joined
- Dec 19, 2005
- Location
- New York
Even though I said I wasnt gonna do it, I have decided to put the E8600 back on water until a used Vapochill LS comes my way, which eventually NOL's autocascade will take the place of the vapo. But for now, I will just have to suffice with water.
The Q820 batch is known by my own research as running slightly hotter than the other chips. I have searched XS forums and found two other members having high temps as well, and guess what the batch was Neweggs Q820. But my VID is very low at 1.156. As Gautam would say, having a low vid and a hotter chip is just the nature of the beast. Switching the Rampage Extreme bios from stock 0306 to 0403 beta has seemed to drop temps about 10*C also for E0 stepping.
One benefit, I am able to run the CPU at .98 vcore at stock frequency 3.33 GHz and 3.60 GHz at 1.01 vcore which would produce ultra low temps around 30*C idle.
For comparison, my first Orthos test for today was 10 X 420 (4.20GHz) at a relatively mild 1.20 vcore. Idle temps are 46*C and loaded temps are right around 62*C which is definitely on the high side, but again its just the nature of having such a low VID from the factory.
Rig Specs:
Asus Rampage Extreme (0403 beta bios)
E8600 Batch Q820A405 VID=1.156v
Cooling = Swiftech Apogee GT WC'ed
Corsair Dominator 2GB, DDR3-1800 (7-7-7-20, 2.0 volt)
Sapphire 4870X2
Screenshot 1 for the day:
She sure does run hot. For some more comparison, I have put my own temperature sensor right underneath a small portion at the top of the CPU underneath the waterblock. Temps seem to idle around 39*C and load around 51*C. Again, that is an external reading between the CPU and the WB and was done for a very short period of time.
Depending on the accuracy of my reading, this possibly tells me it could be a Quality Control production issue with this specific batch Q802, or a core temperature sensor issue. If it is the latter of the two, that means the CPU is actually just reporting the wrong core temperatures and actual temps would be much lower. Again, I am just running my mouth here a bit, dont count on any validity in these statements.
UPDATE.
Just booted windows at 4.30 GHz, at the same vcore of 1.20v. Ran orthos until BSOD and system restart at 1 minute 30 seconds. I currently bumped up vcore to 1.23v and am going to take another shot at 7 minutes of stability or so.
OK. 1.23vcore is good enough for 4.3 GHz and 6 minutes of priming for now. I want to keep pushing before I get any kind of 24/7 stability point. I am really starting to wonder if/when the heat is going to cripple the OC on this processor.
Speak of the devil. Just went downstairs to eat dinner, and figured why not try for 30 minutes of stability or so at this 4.30 GHz and 1.23 vcore. The rig did not hiccup at all for 26 minutes, this is good news for future 24/7 stability point. I also, cranked up the ram to 1720 mhz at stock timings of 7-7-7-20 and 2.0 volts, much closer to its rated spec of 1800 mhz. Did a blend of cpu/ram priority 10 testing and everything looks great. I think I left command rate setting on AUTO, so I am going to manually set 1T on next reboot for the ram.
Bumped up vcore to 1.40 volts for some highest frequency possible testing. Since my vid is only 1.15 , 1.40 is a full .25 points over stock voltage. 1.40 volts was good enough to clock 4.71 GHz and achieve a 9 second Super Pi 1m run. This chip is very nice indeed. As I sit here typing to you at 4.71 GHz I am starting to think that quite possibly the temperature sensors are faulty on this chip.
The Q820 batch is known by my own research as running slightly hotter than the other chips. I have searched XS forums and found two other members having high temps as well, and guess what the batch was Neweggs Q820. But my VID is very low at 1.156. As Gautam would say, having a low vid and a hotter chip is just the nature of the beast. Switching the Rampage Extreme bios from stock 0306 to 0403 beta has seemed to drop temps about 10*C also for E0 stepping.
One benefit, I am able to run the CPU at .98 vcore at stock frequency 3.33 GHz and 3.60 GHz at 1.01 vcore which would produce ultra low temps around 30*C idle.
For comparison, my first Orthos test for today was 10 X 420 (4.20GHz) at a relatively mild 1.20 vcore. Idle temps are 46*C and loaded temps are right around 62*C which is definitely on the high side, but again its just the nature of having such a low VID from the factory.
Rig Specs:
Asus Rampage Extreme (0403 beta bios)
E8600 Batch Q820A405 VID=1.156v
Cooling = Swiftech Apogee GT WC'ed
Corsair Dominator 2GB, DDR3-1800 (7-7-7-20, 2.0 volt)
Sapphire 4870X2
Screenshot 1 for the day:
She sure does run hot. For some more comparison, I have put my own temperature sensor right underneath a small portion at the top of the CPU underneath the waterblock. Temps seem to idle around 39*C and load around 51*C. Again, that is an external reading between the CPU and the WB and was done for a very short period of time.
Depending on the accuracy of my reading, this possibly tells me it could be a Quality Control production issue with this specific batch Q802, or a core temperature sensor issue. If it is the latter of the two, that means the CPU is actually just reporting the wrong core temperatures and actual temps would be much lower. Again, I am just running my mouth here a bit, dont count on any validity in these statements.
UPDATE.
Just booted windows at 4.30 GHz, at the same vcore of 1.20v. Ran orthos until BSOD and system restart at 1 minute 30 seconds. I currently bumped up vcore to 1.23v and am going to take another shot at 7 minutes of stability or so.
OK. 1.23vcore is good enough for 4.3 GHz and 6 minutes of priming for now. I want to keep pushing before I get any kind of 24/7 stability point. I am really starting to wonder if/when the heat is going to cripple the OC on this processor.
Speak of the devil. Just went downstairs to eat dinner, and figured why not try for 30 minutes of stability or so at this 4.30 GHz and 1.23 vcore. The rig did not hiccup at all for 26 minutes, this is good news for future 24/7 stability point. I also, cranked up the ram to 1720 mhz at stock timings of 7-7-7-20 and 2.0 volts, much closer to its rated spec of 1800 mhz. Did a blend of cpu/ram priority 10 testing and everything looks great. I think I left command rate setting on AUTO, so I am going to manually set 1T on next reboot for the ram.
Bumped up vcore to 1.40 volts for some highest frequency possible testing. Since my vid is only 1.15 , 1.40 is a full .25 points over stock voltage. 1.40 volts was good enough to clock 4.71 GHz and achieve a 9 second Super Pi 1m run. This chip is very nice indeed. As I sit here typing to you at 4.71 GHz I am starting to think that quite possibly the temperature sensors are faulty on this chip.
Attachments
Last edited:
