I just built a new system ,AsRock P67 Extreme 4 Gen. 3 MOBO, I7-2700K processor. It will give me a BSOD occasionally. I have a little program (Who Crashed) to read dump files. It tells me Clock Watchdog Timeout error. I have tried to determine just what the problem is. There is a pretty good bit on Google but no answers. Has anyone had to deal with this error. It seems to be fairly common with Wind.7
The blue screen often gives a good clue on what the problem is, often with the driver or file at the end of the text or something in capitals like this MEMORY_ERROR_ " or so. I do not have the same MB as you do, but a ASUS and I have had plenty of BSODs on windows 7 64 bit. First I thought it had something to do with Firefox and flash since BSODs almost every time occured when Firefox was running or was loading a page with flash. But it also happened when the computer was idling some times. I have disabled the "plugincontainer.exe" since it often was spiking CPU. But nothing helped. Did 24 hour memory scan, but nothing there either. And never had BSODs in Linux Mint that I also use. So Windows or some software therein was the problem. Not the hardware. I then disabled the overclocking software (ASUS utilities)that came with the motherboard and I installed and tried some overclocking the first months. Since I didnt really feel it did anything for me I stopped overclocking and defaulted the motherboard to its original values, but let the software be installed and loaded in every boot. Just because it had some monitoring capabilities. Once I stopped their service and removed them from autostart the most BSODs dissapeared. There has been one since I did that but cant locate it. But from having a BSOD every day I have now not had any for two weeks so I think I can say that ASUS utilities was a major part in the BSODs. Maybe you have some motherboard related software in the background?
I had the CPU overclocked to 4.7 on Prime95. So I have backed it off to 4.6 and so far haven't had a BSOD. Right now everything seems to be running great. I did a lot of research on the Clock Watchdog Time out error and there is a lot with Win. 7 64bit Some without any overclocking.
That may likely be the issue. I find that going from 4.6 to 4.7 requires a bit more of a bump in voltage than any increment below that speed.
I really cant see the urgency these days to overclock your CPU just for the sake of doing so, What reasons there are for doing so, just for a little extra push in performance, versus The reasons to just leave it alone, are in favour Against 80/20% Granted everyone would like a blazing fast computer, with minimal price costing, but looking at the facts that are against overclockin, the main reasons not to do it are as follows: CPU: CPU overclocking is the process of running your CPU beyond its listed "clock" speed. When you purchase a processor, the manufacturer has carefully set the CPU at a specific clock speed, such as 2.4GHz or 2,400MHz. That translates to two billion, four hundred million cycles per second. That's an outrageously large number. If each cycle counted a person on earth, you'd be done in less than three seconds. Generally, the higher this number, the faster the CPU because a circuit often completes its task and has to wait for the next clock signal before continuing. By shortening the time between these clock pulses, computer circuits can perform their given task faster. If the clock is set too fast, the circuits will run out of time to finish up their task before another one hits them, and that will lead to system crashes. FSB: The FSB is the main info distribution channel of your PC. It bus-ses data around your motherboard to your CPU, RAM, and all the other peripherals. If you increase the FSB clock speed you will have increased the speed at which the transfer of information occurs. When changing the speed of your FSB, the hard drives and video cards are typically affected the worst and data loss can occur. Most CPUs today run at a preset multiple of the FSB clock speed which is also known as the clock multiplier. Let's use the Intel Core 2 Duo E6600 CPU. It has a x9 (nine times) multiplier which is locked thus cannot be increased. When Intel put your E6600 in the box it was preset at 266 MHz x 9 = 2.4 GHz. However, many overclockers will go into the BIOS and reset this number to 333 MHz x 9 = 3 GHz. By a very simple resetting of the way the computer boots up, you've just increased your system performance by 20%! However, not all overclockers are happy with this level of increased performance. Some will push their E6600 to 400 MHz x 9 = 3.6 GHz which is now a full 50% speed increase over stock. The Intel Core 2 Duo X6800 CPU has a factory preset multiplier of 11. That means that even though the CPU clock is the same as the E6600, the multiplier is 11 instead of nine, providing 266MHz x 11 = 2.93GHz. However, the X6800 has a fully unlocked multiplier, so there are two options for increasing the speed of this system: You can change the 266MHz or the 11 multiplier. The X6800 will usually easily take a multiplier of 13, making it run at 266MHz x 13 = 3.46GHz. Or, you can leave the multiplier as is and increase the frequency to 333MHz x 11 = 3.66GHz. Or get crazy and start experimenting with changing both settings! Video Card: Many of the newer video cards come with software that will let you adjust the clock and processor speed of the card. The adjustments are effectively identical to CPU and FSB as shown above. Voltage: This stressing of your system is brought to even higher levels by the overclocking technique of increasing the voltage. This makes each clock pulse more powerful and less likely to get "lost" in the CPU. Most fried CPUs result from applying a little too much voltage, thus it is heartily recommended that you don't mess with voltage unless you really know what you're doing... and have deep pockets to replace CPUs. Reliability & Stability: All overclocking takes a measure of reliability and stability away from the system. In cases of milder overclocking, this shortening of reliability and stability is negligible and thus not significant. If your system lasts 14 years instead of 15 years it's fairly irrelevant as it will likely be in the junkyard by then, and if you end up with an extra crash every month or so, it's no big deal. When the higher levels of overclocking are reached the reliability and stability decreases exponentially, and some maxed-out systems can only keep running for a few minutes or even seconds at a time. Look at it this way: The manufacturer's preset speeds are the equivalent of a "red line" on a sportscar's tachometer. It is the maximum indicated speed that the CPU should run at to ensure adherence with the manufacturer's warranty. The reasons why a certain clock and multiplier is set are very similar to the reasons for setting an engine's redline: Beyond a particular speed, the CPU/engine might go faster, but at the price of greater heat generation and shortened service life. And just like sportscars, the faster CPUs go, the more likely they are to crash! An electronic phenomenon called electromigration is one of the main reasons for decreased reliability and stability. This phenomenon literally moves particles out of circuits, carving out microscopic canyons and changing the route that the electrons take through the CPU with deleterious results. Furthermore, any increase of speed or voltage of your CPU is asking it to do more work which means more electricity flowing through it which equals more heat. Heat and electromigration are the main reasons why highly overclocked systems fail. Cooling: To keep the heat at a minimum, an overclocked CPU will have to be very well cooled, most likely with a top-end water-cooling system with at least two 120mm fans on the radiators, and three would be even better. Cooling is possibly the single most important factor in overclocking. If you can figure out a way to get your CPU cooler, then generally faster you can overclock it. Best Chill For Your Rig: Air, Water or Peltier Cooling": To OC or not to OC Mild overclocking is generally a safe procedure and will result in noticeable speed increases in the way your PC performs. The problem is what happens when you reach the OC lunatic fringe. There are countless forums on the internet where extreme overclockers compete against each other for bragging rights of who got which CPU up to what outrageous speed. There are no formal prizes or financial rewards in this pursuit, it's just a form of extreme sport which breaks CPUs instead of bones. It's also a very expensive sport as the chase of the ultimate speed leaves behind a trail of sizzled and unusable CPUs. It's not a sport that I would recommend to the faint-hearted or anyone with limited disposable income.
I think we are pretty close to 100% agreement here. Engineers don't design in overclocking features. They design to the design specifications, or to the maximum abilities of the raw materials and current manufacturing techniques (AND as the budget) will allow. Overclocking features are a marketing feature, not a design feature. It is interesting to me (in a scary, not good for consumers, sort of way) that motherboard makers are providing and marketing overclocking capabilities in their boards, aftermarket cooler makers tout their cooling abilities with overclocked system, and even CPUs are being marketed for their OC overhead. Yet Intel, AMD, the motherboard makers nor any of the aftermarket cooler makers will cover ANY damage done to the CPU or the motherboard due to overclocking! This is where we differ a little. I agree completely that electromigration is a real problem - especially in miniature circuits where high current rates exist (like inside powerful ICs, like CPUs). But engineers have known about the affects of electromigration for over a 100 years and assembly techniques and design features are able to compensate (UNDER NORMAL CONDITIONS) for electromigration - at least over the expected lifespan of the product (MTBF stuff). So while electromigration is a real design concern, the greater danger by far is heat and, as noted in my sig, heat is the bane to all electronics. As those electrons pass through circuits, they bang into each other and the conductor walls. Electrons have mass so when they bang into each other, they create friction - wear and tear - and heat. More current means more electrons zipping about, more friction and more heat. Heat begets more heat! In uncontrolled circuits, this cycle of ever increasing heat results in "thermal runaway". Fortunately, designers pay close attention to that. Unless hit with a hammer, it is always heat that takes out a device. Either too much current in the circuit exceeds the capability of the device, or aging (through "normal" wear and tear or accelerated wear and tear due to abnormal - overclocked - voltages or inadequate cooling, or both) weaken the device over time such that it can no longer handle "normal" expectations. Not "possibly" is, but is. Yet sadly, most beginner OCer (and many experienced OCers ) fail to understand the cooling requirements and extra demands placed on the rest of the computer when they OC the CPU. I have seen several water users fry their motherboards because they failed to provide adequate cooling to the heat-sensitive power regulation devices and the chipset components surrounding the CPU socket that depended on the expected air supply from the expected OEM CPU cooler. I have seen CPU sockets that were disformed (warped) due to long term exposure to high temps. The CPU was still fine, but the socket was not. New motherboard time - customer not happy. Also sad is how often OCers OC their main computer - the one they use for work or school projects, on-line banking or other important tasks - then fail to keep a good backup of all their important data. To be sure, there is nothing wrong with being an enthusiast. But do your homework before you take it out on the race track. And be prepared for the worse case scenario - your computer going up in smoke. While rare, it is not that uncommon. So my recommendation is to have at least two computers, one for work, and one for play.
