elPapaGoose

As me ol ma used to say, tis better to have never loved, than to lose and be alone. Not a particularly tender woman...

Bear with me while i tell a tale of woe. Sit and listen, and lend me your ear. Tis a fine night for story tellin'. Grab an ale sire an i will tell you of the one that got away. Which means, let me have an OCD style vent : ) I recently bought the Strix 1070 OC edition. I had to return it for an exchange because of a shot fan controller. Revving. Sudden 100% spin rate. The rpm would read in the 10's of thousands. Was unable to set a custom fan profile. Dud card, so got a new one. My new card performs better (though the fan is still louder than my MSI 980). It still revs at low temps (around 55c) because of what appears to be asus overestimating the fans minimum speed, but with a custom fan profile all is fixed. All of this is business as usual. Got a defective product, returned it for a working one. Amazon customer service is great. Should be end of story. Exceptin for my woes. See, That defective card had attached to it one of the finest pieces of silicon this world ever saw. No ammount of searching on the seven seas, nor high mountains above would produce a treasure so rare. Without me touching a thing, it would boost to 2083MHz and sit there all day. A prize pearl indeed. The white whale. My new card boosts to a mere 2025 (once again at stock settings). 50-60MHz of potential squandered on a poorly produced fan controller. I know i should be grateful for this new card, and what is really a very good performer, but i always return to thinking of the one that got away. My Moby Dick. Imagine the limitless possibilities. The heights we could have soared to, if only given a real chance. Now all life tastes bitter. Tis a fine thing you done. Bearin with a man what is grievin an all. a fine thing. Pardon this old salt his salty tears. My apologies to all who dream of having even 1 of these cards. I am completely aware that i am being a sook.

No replies but i solved it: found helpful info elsewhere. If anyone else has this issue: Turns out that the bios fan control uses a crappy sensor on the mobo. Fans ramp up when it detects power draw increase, but inaccurate readings result in them ramping down again. Fine for non-overclocked systems with simple cooling, but a bit underwhelming for a mobo this high end. Solution: Download fan software that lets you select what sensor CPU fans respond to (which bios wont let you do). Now have the fans listening to the CPU's onboard CPU package sensor and performing as desired. Hope it helps.

Built a new machine. Everything working great and stress tested. Went to set custom fan profiles and ran into an issue. All my fans are PWM controlled and i'm trying to make a profile where they stay constant untill a certain temp then ramp up quickly after that. Went to test the profile and started the stress test, fans ramp up, then very quickly slow down again. They remain at the low setting while aida64 is running, which is a surprise. Not the end of the world, with my fans at 500-600 rpm, 10mins of aida64 stressing, my hottest core peaked at 83c on a hot day in a closed room. 280mm rads and noctua fans FTW. After looking here is the issue: Pwm fans are responding to "CPU" temp reading, not temp core readings (which is normal). The "CPU" temp reading will mostly sit around the temp of my highest core, or thereabouts, but when stressing, after 10 seconds or so the cpu temp reading will drop down again while the temp of the cores continues to climb. When i hit 83c my "CPU" reading was at about 55c or less, so of course the fans are staying at a low rpm. My first thought was that i could just set the fans to ramp up at 50c if the sensor is produving a consisyently low reading, but under normal conditions the sensor is quite accurate (closely mirrors the temp of the cpu cores) and so the fans go crazy everytime i open a program or chrome or what have you. Not the silence i built this thing for. Here is my question: Is this normal behavior for a temp sensor? I understand these things can be out by quite a bit when compared to coretemp, but to be off by 30c for a sustained period of load of 10minutes is wierd to me. It recognises the temp change accurately when the bench starts, so why does the reading go down while temps continue to climb in reality? I can watch this happen with any number of monitoring programs. Does this sound like a faulty sensor? Anyone else have a similar experience? Can someone with a similar maximus viii board replicate this? I would love to avoid dealing with asus customer service to get this fixed. Relevant System specs: Asus maximus viii hero alpha z170, bios 2202 6700k processor H115i cpu cooler with corsair link. (All fan are plugged into mobo).

I reset everything back to stock and opened hwmonitor. Idles sits in the 30's but spiked, even as high as 74 just sitting on the desktop. Does real quick, then goes back down. Not right. I'm going reseat the cooler. I hope its that simple. Have heard of 6700k's having issues with bad internal TIM.

Been overclocking a 6700k and think my temps are too high. During stress tests hit the 80's pretty quick at 1.29 vcore, 4.6ghz. This is with a 280mm h115i cooler. My chip can do 4.7 at 1.35 but temps go up pretty quick. Sound like a contact issue?

So i got irritated and downloaded realbench again, (had done this multiple times) as well as tweaking a few extra bios setting, and at the very least it is runnng. Currently doing a quick 15minute stress test and at least it is running! If only i had changed things one at a time to know which one helped. Hindsight. Testing my chip at 4.6ghz and only 1.27vcore. looks like i have a winner! Hopefully that's problem solved. Thanks all.

Yeah, wish i could. Tried a few things. Set everything back to stock settings. Closed a bunch programs, even turned off HWMonitor. Same crash very early on. Any ideas?

So, i have recently built a system and have been trying to test my oc stability. Realbench stress test very quickly hits an issue with luxmark crashing. This crash is due to my gpu- not my cpu. Even at stock settings. Pretty sure the new bios hasnt fully compensated for the micron memory issue. So i still need to find a way to stability test my cpu overclock. I have aida 64, but want a few more. If you have suggestions, or a secret sauce cure for my realbench woes, let me know! System: 6700k Asus hero alpha z170 32GB corsair ddr4 @3000 Asus 1070 OC edition Samsung 850 evo 500GB Rm850 psu