Hi all, I've been experimenting with basic overclocking and even when I OC just a small amount I get video artifacts in FSX. My 7950 GT came factory overclocked. GPU core is 549 MHz GPU memory is 772 MHz I have the settings in the graphics card set to Application Controlled. All I have done is change the FSB to OC. I locked the PCIe bus at 100 MHz. Processor and Graphics card temps remain within safe ranges after OC. Any help or hints would be most appreciated. Thanks in advance. Craig Lian-Li PC60-B Case Corsair HX Series CMPSU-620HX 620W Power Supply Gigabyte P35 DS3R Intel Core2Duo E6850 @ 3.0 GHz Arctic Cooling Freezer7 Pro Crucual Ballistix 2GB PC2-8500 (DDR2-1066) EVGA 7950 GT OC 512 Mb Plextor PX-810SA/SW-BL 18X SATA DVD Seagate Barracuda 7200.10 500GB 7200 RPM 16MB Cache SATA 3.0Gb/sec Seagate Barracuda 7200.10250 GB 7200 RPM 16MB Cache SATA 3.0Gb/sec Nvidia 163.75 drivers

Bump your Vcore to 1.4v in the bios and verify that voltage level with CPUZ.Download and run Coretemp to monitor CPU temp. These are your individual core temperatures (Tj). These are the temps you want to watch. Verify, but I believe the dual core Tjs are OK well into the 80Cs (for instance, the Q6600/GOs are OK to 100C). Note that this temp is NoT Tc. Your Tc is a different temp (~65C) and only an early warning system for impending Tj problems. Use coretemp to monitor Tj directly.A CPU clock should not cause video artifacts. On first blush, it sounds like that Vcard is 'bout to blow. Send the kids next door before you do these tests. Download the free version of 3DMark06. Use that as your video stress tester. Download and run Prime 95 to test your CPU clock. Well? Is everything still in one piece? Good. Now play Sherlock. Who's doing what to whom? This'll tell the story.

-SNIP->Download and run Coretemp to monitor CPU temp. These are your>individual core temperatures (Tj). These are the temps you>want to watch. Verify, but I believe the dual core Tjs are OK>well into the 80Cs (for instance, the Q6600/GOs are OK to>100C). >>Note that this temp is NoT Tc. Your Tc is a different temp>(~65C) and only an early warning system for impending Tj>problems. Use coretemp to monitor Tj directly.>-SNIP- Tc ( Tcase ) for the Q6600 G0 is 71C MAX. I find that the cores Tj's run typically 7-8C higher. Max disapation is 95 watts. See the following document for details. Table 29 on page 74 is relevant to the Q6600 G0. I wouldn't want to see my core temps above 80C. As it is, overclocked to 3.25Ghz and running Prime95 they stay well below that. ftp://download.intel.com/design/processor...ts/31559205.pdf Paul

More interesting reading.http://www.tomshardware.com/forum/221745-2...mperature-guideSince this is an aviation forum, a powerplant analogy might work. Turbine engines are both thermal and RPM limited. They have EECs (electronic engine controls) that monitor temperatures/RPMs and automatically protect the engine from damage. Intel has adopted the same thermal protection method for the 6 series CPUs. Tj is Intel's version of EGT (exhaust gas temperature). It is the actual temp that is occurring in the turbine section . . . or in this case, in the core. Tc is more of an ambient reading (Tc = Temperature in the Case), and not the actual core temperature. Turbine engines have several temperature designations. One is "Max Continuous." The engine can run at this temp forever. It is designed to successfully deal with thermal environment indefinitely without damage or reduced serviceability (lifespan). If an operator wants to run at a lower temp, that's fine, but it is not saving anything from anything. It's not even prudent precaution. It is though, most certainly, a prudent fuel saving strategy. The Q6600's "max continuous" temp is 95C. This is Tj. Intel designed the CPU to begin to thermal limit this temp. At 100C the CPU shuts down. It is virtually impossible to damage these CPU by thermal means. Remember, Intel builds these primarily for business. They tend to stack these in row upon row of rack after rack in extreme thermal enviroments. A single CPU cradled and coddled under a Thermalright Super-Duper 120 with a bevvy of 120 mm fans blowing away is unknown to most of the CPUs Intel makes. In their more natural business environment, these CPUs Have to take these temps. Actually that 71C Tc measures nothing. It is presented as an artificially low limit to Try to keep the indistros Off their 95C throttling limit. Intel's CPUs would still be fine, but can you imagine the service calls as CPUs throttled off that 95C stop 'round the world? What a mess!Now here's something Really interesting. The 9 series are now using RPM limits.http://www.anandtech.com/cpuchipsets/intel...doc.aspx?i=3184 See if you can get through this. It's dense, but the way forward is clear. Big cooling (water and above) will no longer help. The 9s still have thrermal protection, but are using RPM to limit "work" (and therefore thermal dynamics). The Penryns are hitting a built-in RPM limit Long before they get anywhere close to their thermal limits. The Penryn fanbois need to look a bit closer at their 4G+ O/Cs. Anything above about 4.25G is Not ReaLLy Happening. Hummm . . .