[6850k] Killed my first CPU in 15 years!

[Charlie22911]

Got the 6900k in today and refilled my loop. The first thing that jumped out at me is how much cooler this one runs, and not by a little; we are talking about ~25°C. Prime 95 AVX blend test puts my cores at ~58°C at 4.112GHz 1.26v (So far it's stable). The other chip was doing ~70°C stock and mid-high 80s @ 4.2GHz 1.275v.

Now I know what you are probably thinking, surely application of thermal compound or mounting pressure is to blame?
Well, maybe; but unlikely. When I first built my loop consisting of an EK Supremacy block with a total of 6x120mm in radiator area I was very disappointed with thermal results for the CPU, so I seated the block cleanly with fresh compound... Twice.

Second thing I've noticed, my intermittent cold boot issues are gone completely. I initially suspected the board\bios and attributed these small issues to being on the bleeding edge while assuming they would be corrected in a future BIOS update.
Does this mean I had a dud chip? Honestly I don't know, maybe it was fine stock but crap when run out of spec.

Intel is shipping me a replacement for the 6850k, so now I'm debating whether to sell it or rebuild my aging Thuban based server...

Also, here are some older photos of my current setup for the curious; it was my first custom water loop so a couple of bends are off but not very noticeable. I'll redo them in the future when I next change my fluid.

 

[Justinus]

Got the 6900k in today and refilled my loop. The first thing that jumped out at me is how much cooler this one runs, and not by a little; we are talking about ~25°C. Prime 95 AVX blend test puts my cores at ~58°C at 4.112GHz 1.26v (So far it's stable). The other chip was doing ~70°C stock and mid-high 80s @ 4.2GHz 1.275v.

Now I know what you are probably thinking, surely application of thermal compound or mounting pressure is to blame?
Well, maybe; but unlikely. When I first built my loop consisting of an EK Supremacy block with a total of 6x120mm in radiator area I was very disappointed with thermal results for the CPU, so I seated the block cleanly with fresh compound... Twice.

Second thing I've noticed, my intermittent cold boot issues are gone completely. I initially suspected the board\bios and attributed these small issues to being on the bleeding edge while assuming they would be corrected in a future BIOS update.
Does this mean I had a dud chip? Honestly I don't know, maybe it was fine stock but crap when run out of spec.

Intel is shipping me a replacement for the 6850k, so now I'm debating whether to sell it or rebuild my aging Thuban based server...

Also, here are some older photos of my current setup for the curious; it was my first custom water loop so a couple of bends are off but not very noticeable. I'll redo them in the future when I next change my fluid.

Congrats on the best possible outcome. Replacement of the defective part and a new CPU that is even better than the old one. Glad to see the motherboard is still working well.

 

[BonzaiDuck]

You're saying "your first custom water loop?" And it's a hard-pipe creation? I'd say that either you're really handy at certain things which bode to be tedious and frustrating for me, or you put a lot of time and parts-expense into that rig.

Has anyone built or opened a museum or gallery that displays rigs like this? And if there's a gallery, is there a lucrative market with 2-percenter Wall Street barons who never touched a Phillips screw-driver?

The best I can do so far is a carefully-chosen heatpipe cooler, a CLU-relid-CPU and some green cold-cathode lights. But my 6700K @ 4.6 Ghz shows package temperature maximum of 68C. For several more Franklins for those last 10C degrees, I'll scale back my artistic expectations and see what develops through the next year.

It would be interesting to compare the stress-test peak wattage or OC TDP.

Could you tell me what make and model fans those are, are they PWM, and what is their amperage and top-end RPM/CFM in their spec -- or direct me to a reseller page for the item?

Got the 6900k in today and refilled my loop. The first thing that jumped out at me is how much cooler this one runs, and not by a little; we are talking about ~25°C. Prime 95 AVX blend test puts my cores at ~58°C at 4.112GHz 1.26v (So far it's stable). The other chip was doing ~70°C stock and mid-high 80s @ 4.2GHz 1.275v.

Now I know what you are probably thinking, surely application of thermal compound or mounting pressure is to blame?
Well, maybe; but unlikely. When I first built my loop consisting of an EK Supremacy block with a total of 6x120mm in radiator area I was very disappointed with thermal results for the CPU, so I seated the block cleanly with fresh compound... Twice.

Second thing I've noticed, my intermittent cold boot issues are gone completely. I initially suspected the board\bios and attributed these small issues to being on the bleeding edge while assuming they would be corrected in a future BIOS update.
Does this mean I had a dud chip? Honestly I don't know, maybe it was fine stock but crap when run out of spec.

Intel is shipping me a replacement for the 6850k, so now I'm debating whether to sell it or rebuild my aging Thuban based server...

Also, here are some older photos of my current setup for the curious; it was my first custom water loop so a couple of bends are off but not very noticeable. I'll redo them in the future when I next change my fluid.

 

[2blzd]

my broadwell-e died too. after 10 days of use. I suspect the motherboard took it out when it went down.

So I posted y his on Friday. And my replacement died today. I'm f'n cursed. I really don't make what I'm doing wrong. . .


Is my psu killing my motherboard and cpu?

 

[wingman04]

So far, I've only heard the OP's story firsthand, and Wingman in another thread posted a link to an Intel forum in which a 6600K owner was bemoaning the death of his chip. Purportedly, he only clocked it up to 4.4 Ghz, but there are many reasons this might happen.

If you can't overclock your chip with some reasonable expectations, it takes the fun out of building computers.

So I will mourn, if that is the case.

The question you have to ask your self if folks left the CPUs stock would they still fail? Folks take a chance when they overclock, there can always be something marginal with the transistors or interconnects that pass production, no two CPUs are the same.

When I called Intel RMA for the USA and asked about failure rates of k CPU's she said that she has 3 a day just at here Que. Phone Number: 1-916-377-7000

 

[superstition]

The question you have to ask your self if folks left the CPUs stock would they still fail? Folks take a chance when they overclock, there can always be something marginal with the transistors or interconnects that pass production, no two CPUs are the same.

I've read reports of boards overvolting CPUs at stock, though.

 

[bigboxes]

Thanks, that's good to know. I'll have to drain the loop (ugh) and pull the CPU when I get home from work and start that process.

You need to drain the loop to get heatsink off? Are your tubes not flexible? Very impressive rig, regardless.

 

[BonzaiDuck]

I've read reports of boards overvolting CPUs at stock, though.

Well, both Wingman and I know we're taking a risk -- he's running 6600K at 4.5 and my 6700K is at 4.6.

I think there's enough truth about the boards and their default settings. I wouldn't know about the Broadwell E processors. If it's not like comparing apples to oranges, then it's a comparison of Valencias to Navels. For that analogy, I can't tell the difference in the juice.

Are the processors dying from too much or too little juice? Seems more likely, if our colleague has built a phenomenal cooling system like that.

Between the boards and my OC settings, both VID and VCORE max out in values close to the stock auto board settings.

 

[DrMrLordX]

I don't think you're going to kill a CPU from giving it too little voltage/current. Er, juice.

 

[Charlie22911]

Not sure if this would be considered a thread necro, but I wanted to provide an update. CPU is still with intel waiting on RMA, I have a box put together to house it for unRAID purposes (6x4TB WD Reds w/ 2x old intel 180GB SSDs for cache, 32GB ram, Fractal Define R5).
The 6900k that has replaced it has been doing great, final settings I settled on for 24/7 use are 8 hours linpack 64bit AVX stable w/90% ram allocation:

4GHz core @ 1.24v core \ 1.92v VRM
2.8GHz ring
3GHz Ram

Spoiler: OCCT Graphs

screen shot on pc

You're saying "your first custom water loop?" And it's a hard-pipe creation? I'd say that either you're really handy at certain things which bode to be tedious and frustrating for me, or you put a lot of time and parts-expense into that rig.

Has anyone built or opened a museum or gallery that displays rigs like this? And if there's a gallery, is there a lucrative market with 2-percenter Wall Street barons who never touched a Phillips screw-driver?

The best I can do so far is a carefully-chosen heatpipe cooler, a CLU-relid-CPU and some green cold-cathode lights. But my 6700K @ 4.6 Ghz shows package temperature maximum of 68C. For several more Franklins for those last 10C degrees, I'll scale back my artistic expectations and see what develops through the next year.

It would be interesting to compare the stress-test peak wattage or OC TDP.

Could you tell me what make and model fans those are, are they PWM, and what is their amperage and top-end RPM/CFM in their spec -- or direct me to a reseller page for the item?

I watched a bunch of videos on YouTube of how people were bending their tubing, which helped considerably. It wasn't difficult though, I probably went through 6-7ft of tubing to get these bends but once you get a feel for how the material responds to heat it gets quite easy.
I must revise my thermal statements however, I was looking at the debug display for my thermals which as it turns out displays core temperature and not package temperature; see above graphs. Peak package TDP according to aida64 is ~180w during prime95 small FFT.

The fans I'm using are RGB Thermaltake RIING 120mm PWM fans, which are rated for 40.2 CFM:
http://www.newegg.com/Product/Produ...=thermaltake_riing_rgb-_-35-106-355-_-Product

I like them quite a bit, the loudest thing in my rig by a fair margin are the WD Black mechanical drives.

You need to drain the loop to get heatsink off? Are your tubes not flexible? Very impressive rig, regardless.

Unfortunately I have to drain the loop do do anything involving the GPUs or CPU as I am using PETG hard tubing. It's not really difficult to do but I'm lazy and don't like doing extra work haha.