X99-WS 10G OC (non-thermal) shutdowns

[foolishlywise]

After doing some more digging with a (badly binned) 6800k mentioned in a previous thread, I'm sticking to a more sane 4.1GHz at 1.265v. This seemed stable until the PC just shut down.

No BSOD. No event log entry apart from 'The previous system shutdown was unexpected', no error on RealBench. It just shut down point blank. Heard the usual PSU click and yep gone. As far as I've read, a bad OC will produce a blue screen. This issue isn't - thermals stay below 60degC and the PSU was good in the previous build, never skipped a beat.

 

Trying RealBench on stock now (everything set to auto in bios) and it seems to be going alone fine (need another hour to see). Could even the slightest OC be causing the pc to just turn off? Or am I looking at a MB issue? 

 

EDIT:

 

System spec:

6800k

X99 WS 10G

24GB Vengeance LPX 3000Mhz 8x2 4x2 

Custom loop (D5, 1x 140mm, 1x 240mm, cpu in loop only). 

[knightslugger]

Having the same issue with the build below (different board, i know). NO warning, no blue screen, no system events. Currently waiting for it to happen again without the GPU installed. Mine shuts down under several hours of idle conditions though and attempts to restart (probably a BIOS power outage condition) but fails POST with a BOOT_DEVICE_LED on the MB. a manual powerdown powerup resolves the issue... until it happens again after several hours of idle. Could be a MB issue, could be a GPU issue. so far it's not a HDD/SSD issue (switched those out) and not a PCIe Wifi adapter issue (removed and it still powered down) and it's not a RAM issue (MemTest86 passed with flying colors). If it happens again, i'm sending the board in. if it doesn't, i'll install the GPU into a different PCI slot and see if that makes it drop out. if it does, i'll send the GPU in.

 

so far, no problems. It's an adventure every time i wake up or get home from work. will it be on when i get back? will it be on when i wake up?

[foolishlywise]

17 hours ago, knightslugger said:

will it be on when i wake up?

I feel you! Let me know if you get this sorted - I think mine was due to a bad overclock or something - switching to stock voltages and multiplier sorted it out, however I haven't had enough time to test (need to leave overnight with nothing urgent going on). 

[done12many2]

Are either of you guys overclocking CPU cache?  Most instant shutdowns without warnings are a result of cache related issues.

[knightslugger]

*deleted*

[knightslugger]

1 hour ago, done12many2 said:

Are either of you guys overclocking CPU cache?  Most instant shutdowns without warnings are a result of cache related issues.

I've not touched my cache speed and my shut downs occur weather or not i'm OC'd. So far though, my system remains online. nearly 72 hours straight. 

[foolishlywise]

5 hours ago, done12many2 said:

Are either of you guys overclocking CPU cache?

Haven't touched the cache when OC'ing. Literally, a voltage is set with multiplier - that's it

[done12many2]

5 hours ago, knightslugger said:

I've not touched my cache speed and my shut downs occur weather or not i'm OC'd. So far though, my system remains online. nearly 72 hours straight. 

1 hour ago, FoolishlyWise said:

Haven't touched the cache when OC'ing. Literally, a voltage is set with multiplier - that's it

 

The second most frequent cause for immediate shutdowns after cache issues is Gremlins.  

 

Good luck in tracking down the issues gents.

 

 

 

[foolishlywise]

On 15/03/2017 at 7:27 PM, done12many2 said:

The second most frequent cause for immediate shutdowns after cache issues is Gremlins.  

I think I've managed to prove why mine is shutting down. Motherboard. Left everything stock speed (auto voltage and all) and came home to see the PC had uncleanly shut down three times in the space of around six or seven hours now. Might just ditch the WS board and go with the X99 Deluxe II. Idk. 

 

@knightslugger Any luck with you bro?

[knightslugger]

10 minutes ago, FoolishlyWise said:

I think I've managed to prove why mine is shutting down. Motherboard. Left everything stock speed (auto voltage and all) and came home to see the PC had uncleanly shut down three times in the space of around six or seven hours now. Might just ditch the WS board and go with the X99 Deluxe II. Idk. 

 

@knightslugger Any luck with you bro?

well my board stay'd lit up for over 72 hours. By now i should have seen SOMETHING. so I plugged everything in where it was before EXCEPT the Video card. I put that in the second slot just to see if it would cause a sudden shutdown. OC'd everything like it was. 1.36v Adaptive, 46x100BCLK, 3200XMP, 2100MHz on the GPU. Solid stability on all fronts.

 

so far, so good. i have to give it the whole weekend though.

[knightslugger]

Computer pooped out around 48 hours.

 

RMA'd the GPU. Stuck a GT610 in the same slot as before. 24 hours in, no shutdown.

[done12many2]

4 minutes ago, knightslugger said:

Computer pooped out around 48 hours.

 

RMA'd the GPU. Stuck a GT610 in the same slot as before. 24 hours in, no shutdown.

 

The post before your last said you made it 72 hours on the last configuration so give it another 48 or more hours. 

[knightslugger]

Just now, done12many2 said:

 

The post before your last said you made it 72 hours on the last configuration so give it another 48 or more hours. 

that's the plan! run it til it surpasses previous benchmarks! if it still poops out, the board goes back.

[knightslugger]

@FoolishlyWise @done12many2

 

well she didn't make it to 48 hours sadly. Still powers off with either GPU in the slot.

 

MB is going back to Asus. If they can't figure it out, I'm moving on to an AMD build.

[done12many2]

14 minutes ago, knightslugger said:

@FoolishlyWise @done12many2

 

well she didn't make it to 48 hours sadly. Still powers off with either GPU in the slot.

 

MB is going back to Asus. If they can't figure it out, I'm moving on to an AMD build.

 

That sucks to hear man.  If you can hold out a bit longer, x299 is going to drop in 5 or 6 months.  Ryzen is a solid choice as well.

[knightslugger]

5 hours ago, done12many2 said:

 

That sucks to hear man.  If you can hold out a bit longer, x299 is going to drop in 5 or 6 months.  Ryzen is a solid choice as well.

I just don't think i'll be able to afford them.

[done12many2]

25 minutes ago, knightslugger said:

I just don't think i'll be able to afford them.

 

Do you have a specific need for more cores?  Otherwise, I'd continue to troubleshoot your current setup and stick with your z170/6700k setup.  

[knightslugger]

2 hours ago, done12many2 said:

 

Do you have a specific need for more cores?  Otherwise, I'd continue to troubleshoot your current setup and stick with your z170/6700k setup.  

it would be nice to have more for rendering summer time on-road / off-road truck and motorcycle racing and PRS competition footage. I cover several different disciplines and my summer is already chalk full on weekends. I think between beginning of May and the end of September i have 4 weekends off? maybe?

[foolishlywise]

21 hours ago, knightslugger said:

well she didn't make it to 48 hours sadly. Still powers off with either GPU in the slot.

Ah man. Life is a through bitch sometimes. 

 

Like @done12many2 says, stick with that Skylake for now if funding is an issue - I'd say its no slouch for rendering. Although I'd lean toward it being something up with the motherboard (or dare I say, CPU prehaps(??) if it still screws up after the mobo RMA). 

 

Mine hasn't rebooted after being left at stock for a few days it seems (touch wood) but, as is the nature of my luck with anything, it'll probably go as soon as I start to believe its okay. 

[dexT]

Haswell-E is good for overclocking but Broadwell-E is a dog. I'd go as far to say Haswell-E is the faster CPU even though it has lower IPC

 

Maybe you guys should try a 5820K. The X99 WS boards are really nice and should OC really well, my M does anyway.

[knightslugger]

5820K 6 core @ 3.3GHz for $390 + $300 for a mainstream board

or

$370 for a 8 Core R7 1700X @ 3.4GHz + $170 for a mainstream board.

 

While the 5820K at a good 4.5GHz OC gets a CB15 score in the 1500's, the 1700X at stock speeds starts at 1500, and i've see the OC profiles at 4.0GHz with faster memory garner a good 1700+ and honestly gaming at 4k60 isn't going to punish the CPU all that much to start with. hell if anything i'd benefit from SLI. might as well save the money and get another 1070FTW. then maybe CPU would start to suffer...

[dexT]

22 minutes ago, knightslugger said:

5820K 6 core @ 3.3GHz for $390 + $300 for a mainstream board

or

$370 for a 8 Core R7 1700X @ 3.4GHz + $170 for a mainstream board.

 

While the 5820K at a good 4.5GHz OC gets a CB15 score in the 1500's, the 1700X at stock speeds starts at 1500, and i've see the OC profiles at 4.0GHz with faster memory garner a good 1700+ and honestly gaming at 4k60 isn't going to punish the CPU all that much to start with. hell if anything i'd benefit from SLI. might as well save the money and get another 1070FTW. then maybe CPU would start to suffer...

If you still have an X99 board a used 5820K is less than $300. Do whatever you want though, as I really don't care.. this was just a suggestion. And mainly for the OP.

 

[dexT]

On 3/14/2017 at 10:29 AM, FoolishlyWise said:

After doing some more digging with a (badly binned) 6800k mentioned in a previous thread, I'm sticking to a more sane 4.1GHz at 1.265v. This seemed stable until the PC just shut down.

No BSOD. No event log entry apart from 'The previous system shutdown was unexpected', no error on RealBench. It just shut down point blank. Heard the usual PSU click and yep gone. As far as I've read, a bad OC will produce a blue screen. This issue isn't - thermals stay below 60degC and the PSU was good in the previous build, never skipped a beat.

 

Trying RealBench on stock now (everything set to auto in bios) and it seems to be going alone fine (need another hour to see). Could even the slightest OC be causing the pc to just turn off? Or am I looking at a MB issue? 

 

EDIT:

 

System spec:

6800k

X99 WS 10G

24GB Vengeance LPX 3000Mhz 8x2 4x2 

Custom loop (D5, 1x 140mm, 1x 240mm, cpu in loop only). 

What power supply? X99 chips draw a lot of power when overclocked.

[knightslugger]

3 hours ago, FoolishlyWise said:

Ah man. Life is a through bitch sometimes. 

 

Like @done12many2 says, stick with that Skylake for now if funding is an issue - I'd say its no slouch for rendering. Although I'd lean toward it being something up with the motherboard (or dare I say, CPU prehaps(??) if it still screws up after the mobo RMA). 

 

Mine hasn't rebooted after being left at stock for a few days it seems (touch wood) but, as is the nature of my luck with anything, it'll probably go as soon as I start to believe its okay. 

well if it still happens after the MB swap, i'll be absolutely certain to look at a different CPU. The thing is that it was doing this even before i pealed the hood back, so it wasn't anything that I did. Convincing Intel of that will be tough though. if they tell me to piss up a rope, I guess that's fine.

 

I did order an air cooler to see if maybe MAYBE the AIO pump power is falling on it's face and its thermal-protecting? you'd think you'd see a warning screen for that, not a lock up on reboot as BOOT_DEVICE error, only to re-reboot and everything be fine. I admit, this is a long shot.

 

Hope your system stays rollin' though.

[foolishlywise]

On 2017-3-27 at 2:38 PM, dexT said:

should try a 5820K. The X99 WS boards are really nice and should OC really well,

On 2017-3-27 at 4:48 PM, dexT said:

What power supply? X99 chips draw a lot of power when overclocked.

 

Got myself a HX1000i. I built someone a system once which needed the 1000w and I noticed no coil whine from the PSU so decided on that. 

 

I'm feeling a 5820k - Amazon UK are total out of stock on it though. I can't keep the 6800k stable at 4.2 anything below 1.343v.