Unable to pass IBT with maximum score despite using default settings.
Unable to pass IBT with maximum score despite using default settings.
Well, there was quite a puzzling occurrence when I tried again with a stable 4.5 ghz OC. I had completed ten runs on both standard and high settings at a vcore of 1.38, then gradually reduced the vcore, retested, lowered it once more, and tested again. The last adjustment caught me off guard. In the "Normal" column, the recommended vcore was 1.355v, but since I was performing well at 1.36, I went with 1.34 and lowered the LLC from 7 to 6. Surprisingly, this actually improved stability. It completed all ten high-voltage passes smoothly, which was better than before when using higher voltages. On very high settings, it was failing on pass 2 or 3, but lowering the voltage helped. This result is really unexpected for me. Adjusting the OC is tricky to grasp. I’m not sure if the lower LLC contributed too. This table I found has been really helpful. http://forum.giga-byte.co.uk/index.php?topic=7760.0. Just wanted to share my temporary success, hoping it doesn’t lead to more confusing outcomes and that I won’t fall behind on standard tests anymore, for some unknown reason. Murpheys Law applies here.
Your cooling performance isn't the best, it barely matches a Noctua NH-D14, and the D14 with 120mm 110cfm fans will surpass it. The main factor for successful overclocking is maintaining adequate CPU temperature to support the boost. Just start IBT on standard settings—it's just the initial step toward finding stability. The advice provided here applies to an ASRock motherboard but can still be applied to your GigaByte board. You can find more details at the following links: With the Sandy Bridge CPU, your memory performance might be limiting you—please specify your memory configuration (16G 2x8G or 4x4G) for a more accurate guide.
4 Ryan6 shared his thoughts on the matter. He believes his current cooling setup isn't optimal, especially when compared to a Noctua NH-D14, and suggests using the 120mm 110cfm fans from D14 would be more effective. He emphasizes that maintaining CPU temperature is crucial for successful overclocking. He recommends starting with standard BIOS tuning before moving on to more advanced techniques. The advice he provided was originally intended for an ASRock motherboard but could still apply to your GigaByte board.
He linked several guides that discuss overclocking stability, including one about Sandy Bridge CPUs and another about memory speed limitations. He also mentioned the importance of checking his RAM specifications—whether it's 16G 2x8G or 4x4G—and shared a link to a detailed guide on reaching CPU overclock stability.
Regarding cooling, he agreed with the sentiment that his current system isn't performing as expected. He noted the lack of clear information online about the unit and expressed skepticism about its quality, especially since it came pre-installed in his first setup. Despite this, he acknowledged learning a lot from his experience and is now more cautious about purchasing pre-built desktops.
He plans to invest in a new liquid cooler for his rig due to persistent overheating issues at higher speeds. He remains surprised to find a stable 4.5GHz build today, though he remains cautious given his recent experiences. He remains open to suggestions on suitable liquid coolers that fit his current hardware.
The issue is widely recognized and might be impacting your IBT tests. Refer to the article at If you're running 64c under heavy load, your temperature readings may not reflect the actual performance. Maya and Mudbox can utilize video cards for calculations, which significantly speeds up processing compared to relying solely on the processor. Ensure these programs are set to use the available video cards. This adjustment will reduce rendering times considerably and also ease the processor's workload. The improvement in stability when lowering voltages stems from internal leakage within the processor. Not all processors handle higher voltages equally; some can withstand more without overheating. Electron leakage between components generates hot spots, leading to instability. This effect may not always appear on temperature monitors, depending on where and how long the leakage occurs. As clutchc mentioned, I would think about upgrading to a better PSU.
The issue is well documented and may impact your IBT tests. Refer to the article at If you're running 64C under heavy load, you should check your temperatures. Maya and Mudbox can utilize video cards for calculations, which is significantly faster than relying solely on the processor. Ensure these programs are set to use the available video cards to reduce rendering times and lessen processor strain. The improvement in stability when lowering voltages stems from internal leakage within the processor. Not all processors handle higher voltages equally; some can tolerate more without issues. Electron leakage between components creates hot spots, leading to instability. This effect may not always appear in temperature readings, depending on where the leakage occurs and how long it takes for errors to manifest. As clutchc mentioned, I’d recommend upgrading to a higher-quality PSU if applicable. He’s using a Sandy Bridge CPU, not Skylake, so your comment is quite off-topic. Your situation is unrelated. His 2600K should handle a 4.5ghz overclock with its current limitations, provided he applies the correct settings and has sufficient cooling.
I am definitely willing to hear suggestions for new liquid coolers that match this case/socket configuration. Thank you for the details. I will follow up tomorrow. Please note my recommendation: use the info you shared and test how much performance you can achieve before thinking about a replacement cooler, since your current cooler should handle at least a 4.5ghz overclock with the right BIOS settings.
Make sure your homework is done before you comment. The i7-6700k is only available in the Skylake version, not Sandy Bridge. He also shared the incorrect socket, and it's 1151, not 1155. 1155 is for Ivy Bridge, while the I7-3770k uses 1150, similar to the i7-4790k I have. Considering this, he does have a processor issue, which might be causing the errors before it completes multiple passes in IBT. You can search for news about it everywhere. Sorry, but it seems to be what it is.
Indsup states that Sandy Bridge supports 1150, similar to the i7-4790k model currently in use. It also mentions that Sandy Bridge utilizes the 1155 socket.
You should complete your assignments first before commenting. The i7-6700k is available only in the Skylake version. It doesn’t match the Sandy Bridge or 1151 socket. 1155 is for Ivy Bridge models, while the I7-3770k uses 1150. Given this, it seems he has a processor issue that might be leading to errors during testing. Searching for updates online is worth a try. Sorry but this is the current situation.
4Ryan6 :
rthughes :
I am most definitely open to suggestions as to new liquid coolers I could get that will fit with this case/socket set.
Thanks again for the info. Ill report back tomorrow
😉
I suggest using the information I've provided you and see how far you can get before even considering a new cooler, as the cooler you have should at least support a 4.5ghz overclock if you use the correct BIOS settings.
Please list your exact memory model #s they are on a sticker on the side of the modules, list model#, speed, voltage, and the timings listed?
gahh I am such an idiot. I just accidentally hit solved on this post. Dont get me wrong, you have been incredibly helpful, I just wasn't done getting help lol. Oh well, apparently there is no way to undo it. Here is the exact memory I have. http://www.corsair.com/en-us/vengeance-1...m4a1600c9b.
Also, another funny example of my apparent idiocy. I was looking in my case earlier at the fans mounted on my radiator and I realized that one of them was set up as an out fan and one of them was an in fan. Which of course is highly counterproductive as the two top case fans and one of the radiator fans are trying to pull air out of the radiator and the case through the top vents, while one of the radiator fans is sucking the air into the case. I can only assume that last year when I completely took my entire rig apart in order to actually learn in some practical detail how the parts were laid out, what cords did what, etc, and also to mount in a few new hard drives, that when I was putting it back together I must have put one of the fans back on wrong and never noticed. I remember being up still at like 430 am trying to finish putting it back together because I was crazy determined but also mad sleep deprived lol. At least I noticed it and I fixed it now. That should help out a little.