Parsleybravo Posted September 21, 2009 Posted September 21, 2009 Okay, so I got my new system built and things were going swimmingly until I started getting random reboots. It happens after I've been running a game for an hour or so. Here's what I'm working with: EVGA X58 SLI Classified i7 975 @ 3.49 12GB OCZ Gold @ 1333 GTX 260 (x2) X3 Modular 1000W PSU Windows 7 Ultimate 64-bit On the cooling side of things, I'm running three, 23-cm fans (front, side, top) and one, 12-cm at the back. My CPU is cooled by a Thermaltake 760i. I really have no idea why it's rebooting like this. The first time it happened, I thought something might have overheated, but I immediately popped into the BIOS to check the "System Health." Everything seemed fine. I'm new to overclocking, but the only changes I've made were enabling "Dummy OC" in the BIOS and upped my RAM frequency to 1600. After the second reboot, I dropped the RAM back to 1333, but no joy. I've attached some screens of the info coming out of E-LEET: If anything looks a little off, or if you guys have any other suggestions, they would be most appreciated. Share this post Link to post Share on other sites More sharing options...
xchrissypoox Posted September 21, 2009 Posted September 21, 2009 Okay so I had this problem and it was that my CPU VTT and DRAM Voltage were to high and were overheating or something (Someone smarter than me is it the NB?) I stopped having crashes with mine set at 1.20v and 1.50v respectively. Lower the voltages one at a time in small increments and run memtest+ test 5 for 20 passes and then continue lowering, once you fail increase the voltage and fun like 2 hours of test 5, and 2 hours of all tests. Anyway I'm pretty sure cpu vtt and dram voltage increase the NB temp and the NB is over heating so put a fan on it and fine tune your voltages. (If I'm wrong please correct me, I don't have time to google it cause I gotta go to class, so don't "let me google that for you" me please ) Share this post Link to post Share on other sites More sharing options...
Mikesnow Posted September 21, 2009 Posted September 21, 2009 It could be a few things tbh, if you can rule out thermal for sure, make sure you have a good bios, you might want to flash a new bios to it. There were some issues with thier smart fan settings or sumthing. I vaguely recall some issues with some of the bios, sleep mode settings needing fixed a few versions back as well. Not many folks use the dummy oc thing, if you go check out evga's forums they have some bios templates you can copy to see if that helps as well. Share this post Link to post Share on other sites More sharing options...
Crazy_Nate Posted September 21, 2009 Posted September 21, 2009 How about a baseline? What does it do at stock settings? Make sure that the memory voltage is set to the manufacturer recommended VDIMM. Then run programs like: memtest86+, Prime95, OCCT (and its PSU test), 3DMark, etc. It's often best to determine if the hardware is working properly, before you start to overclock. Share this post Link to post Share on other sites More sharing options...
ccokeman Posted September 21, 2009 Posted September 21, 2009 I doubt that the 760i will handle the thermal load of an I7 overclocked much less at stock. Also make sure the chipset is cool. Take a screen shot of the CPU temperatures under load with realtemp. THe QPI volts need to be upped when you are pushing the memory clock speeds as well Share this post Link to post Share on other sites More sharing options...
Parsleybravo Posted September 22, 2009 Posted September 22, 2009 Okay so I had this problem and it was that my CPU VTT and DRAM Voltage were to high and were overheating or something (Someone smarter than me is it the NB?) I stopped having crashes with mine set at 1.20v and 1.50v respectively. Lower the voltages one at a time in small increments and run memtest+ test 5 for 20 passes and then continue lowering, once you fail increase the voltage and fun like 2 hours of test 5, and 2 hours of all tests. Would those voltages be specific to the MOBO only, or would they be affected if you and I have different processors/RAM? It could be a few things tbh, if you can rule out thermal for sure, make sure you have a good bios, you might want to flash a new bios to it. There were some issues with thier smart fan settings or sumthing. I vaguely recall some issues with some of the bios, sleep mode settings needing fixed a few versions back as well. Not many folks use the dummy oc thing, if you go check out evga's forums they have some bios templates you can copy to see if that helps as well. I'll try updating the BIOS. The link to S61A on EVGA's website is broken at the moment, so that might have to wait. How about a baseline? What does it do at stock settings? Make sure that the memory voltage is set to the manufacturer recommended VDIMM. Then run programs like: memtest86+, Prime95, OCCT (and its PSU test), 3DMark, etc. On a stock setup, things seem to work fine. I ran Memetest86+ for seven hours before I tried overclocking anything, and it had no errors. And, according the the documentation for my RAM, 1.65 is the recommended voltage. Perhaps that rules out a problem with the RAM? I doubt that the 760i will handle the thermal load of an I7 overclocked much less at stock. Also make sure the chipset is cool. Take a screen shot of the CPU temperatures under load with realtemp. THe QPI volts need to be upped when you are pushing the memory clock speeds as well Stock, my CPU was running 29C idle, 34C load. With the "Dummy OC" enabled, it was running 32C idle and 42C-44C under load. Unless I miss my mark (which is entirely possible), those are pretty reasonable temps for a 975, no? As far as the QPI voltage, is that done through incremental trial-and-error, or is there a rule of thumb? Thanks for the replies, guys, I appreciate the insights. Keep 'em coming! Share this post Link to post Share on other sites More sharing options...
Zertz Posted September 22, 2009 Posted September 22, 2009 Use CoreTemp or RealTemp to monitor core temperature. E-LEET only gives processor temperature which is always much lower than the cores actually are. Share this post Link to post Share on other sites More sharing options...
kitfit1 Posted September 22, 2009 Posted September 22, 2009 (edited) With the Classified, it's well known that it overvolts the ram like crazy. With mine, for instance, 1.65v set in the bios is in fact 1.68v when measured from the measuring pad. To get 1.65v i set it to 1.63v in the bios. Forget about any readings in the monitoring section of E-Leet as well, they are ALL wrong. You also don't need 1.4VTT for that ram speed and timmings, set it to +25mv. Edited September 22, 2009 by kitfit1 Share this post Link to post Share on other sites More sharing options...
Recommended Posts
Please sign in to comment
You will be able to leave a comment after signing in
Sign In Now