Jump to content
Search In
  • More options...
Find results that contain...
Find results in...

N3xiz

Member
  • Content Count

    5
  • Joined

  • Last visited

Awards


This user doesn't have any awards

About N3xiz

  • Title
    Newbie
  1. Solution: After replacing the RAM the problem was still there and the same Core kept failing in any stress test when using FPU in both Aida64 and Prime95 newest versions. I then RMA'd the CPU and replaced it, after 40 minutes of Small FFT and 15 minutes Large FFT there were no more errors, I know this is a rather short time but it failed instantly before, I also enabled Round-off checking and SUM check which caused the faulty CPU to fail instantly on one core as I said. So after all it was the FPU of Core #1 which got somehow damaged, pretty strange to me but I always have bad luck...
  2. Update: So I played around a bit and tried setting my multiplier to 30 so it would run with 3,0Ghz on all cores instead of 3,6Ghz and 3,8Ghz on the Turbo Core and set the VCore to 1,2 just to be safe auto would not mess up things, with that settings I surpassed everything I could not before, it did blend with 400k which is the first test without errors instead of stopping workers 3 and 4 immediately, funny enough this core namely core #1 should be my turbo core according to bios but the two threads of core #1 stop always first... I then proceeded to set everything on optimized defaults again without changing voltage and again Core #1 was the only one failing but this time with worker #4 after 2 minutes and worker #3 after 6 minutes, this seems strange to me because core #1 should be the chosen one to outperform them all but fails first. Does this sound like something ram related? because I can't imagine a processor to sometimes work and sometimes not given the same circumstances, again, before it failed immediately with default settings. PS: I only used one ram stick for all my testing
  3. The Core Voltage rises slightly under AVX usually about 0.01 or something but I really feel like this should not be an issue with all stock settings and optimized defaults, so all is set like Intel and Asus intended. I used the most recent Prime95 version, literally downloaded it yesterday.
  4. I already used the Preset Small FFT, Voltage is on stock setings, 1,1V which worked about a year ago, it fails instantly when starting or one core first and the second shortly after, temps do slightly touch 70C.
  5. Hey folks, I've been browsing the internet for days now and I am afraid my CPU may has gone bad. I recently had Overclocked my i7 6800K to 4,1GHz on 1,27VCore, all fine and max temp in prime95 was 85C'ish (on air), then I wanted to test stability and used blend test but it failed after 2-3 minutes with a rounding error, so I decided to downclock but the problem persisted, I thought it may be the ram so I loaded Optimized Defaults and tried again, it still kept failing and ALWAYS on the same two cores and now instantly when the test starts, I tried only using one ram stick but for all four I got the same result. I then ran Intel Processor Diagnostic Tool which says it was all fine but blend and small fft kept failing the same two cores... I really don't know what to do because I remember testing with prime95 for max temps with that cooler and nothing failed, this was on stock too. One thought was that the oc settings applied by the mobo had fried the fpu in that two cores, I left everything on auto except of VCore and Cache Voltage because I overclocked that too but never went above 1,3V on VCore and never above 1.2'ish on cache. I also tried the overclock preset when starting to oc because I thought that would be safe but it bumped Core to 1,5V and it didn't even boot so I did it manually with fully manual mode enabled still everything on auto except VCore and Cache. Can it be that the fpu on that two cores went bad because of oc? With one stick it managed about a minute before the workers stopped but its always the same two, I will try different ram tommorow because I ordered some trident z rgb ram but small fft shouldn't fail because of ram or can it? I also tried prime95 on a ubuntu live usb stick but nothing changed. I would be really thankful for replies, greetings from germany ^^ System Specs: Motherboard: Asus Strix X99 Gaming BIOS 1801 CPU: i7 6800K RAM: Kingston HyperX Predator 3333Mhz PSU: be quiet! Dark Power Pro11 Prime95 newest version PS: Also tried Intel Processor Diagnostic Tools Burn In Test which lasts 120 minutes and it passed Math, Prime Number and Floating Point, shouldn't an Intel Tool be reliable?
×