Setecastronomy

*Disclaimer, I'm pretty new to overclocking, be it CPU or GPU, trying to learn as much as I can, so I don't have much experience with this stuff, just wanted some feedback on this. So I've got a 6700k in a Maximus VIII Hero, with 32gb Corsair ddr4 lpx ram 2400 mhz. CPU is cooled by a h110igt. I've had this running for about a month, and up until now all i've done as far as 'over-clocking' is turning on the XMP profile in the bios. I wasn't sure if I wanted to overclock manually, in the bios, or use the AI Suite 3 auto-overclocking utility. I'm looking at the following two videos as a reference: Manual overclock: This one seems more intimidating, but she's using the same cpu/motherboard combo I am, and she details EXACTLY what to type in to get the desired overclock. Auto overclock: He explains basically how the utility works, but doesn't show what settings he entered before he started the auto-overclock utility. I flipped a coin and decided to start with the auto overclock option first, just to try it out. I set the parameters so that the voltage shouldn't exceed 1.35v and temp shouldn't exceed 80C during the process and then let it do it's thing. It got up to 4.8 ghz, then gave me a BSOD and restarted and finally gave me a stable overclock of 4.7 ghz. After that was done I ran RealBench for 30 min and while that was running I had HWmonitor up to keep an eye on temps and voltage. I noticed that during the Real Bench stress test, temps were sitting around 75-80C, CPU at 4.7 ghz and 100% load, BUT, the voltage was sitting consistently at or around 1.25V... So my question is, should I take this to mean my 6700k is running stable at 4.7ghz at 1.25V? Reading through various forums, including this one, I've seen most people requiring 1.35v to get 4.7ghz stable on this chip... Like I said, I'm still really new to this stuff, so am I misinterpreting something here, or did I just get lucky and get a really nice chip? I'm really curious now to try the manual overclocking (through the bios) method that Joanne showed in her video and see how far I can push it. Just curious on what someone with more experience with this thinks. Should I stick with the overclock I've got now or would it be more stable/safe to manually overclock through the bios?

Oh I'm not mad at EK about it, don't get me wrong. On a long enough time line, any company is going to have a faulty product go out the door. This was my first time buying from EK, and I got burned, but that's just my bad luck. I still think EKWB is an awesome company, they have a great reputation in the industry for a reason. I mean look at the way they have announced this recall, I really respect the way they aren't trying to hide anything that's going on. Yes, in my specific case I've had a difficult time coordinating this whole return shipping/refund thing, it's been almost three weeks since I first opened the RMA, and I still haven't had the defective unit shipped back yet, but I'm not angry about it. I get that the company is really busy dealing with this, and they are half way around the world from me. This incident has scared me away from liquid cooling for a bit, but if/when I decide to go back to doing a custom loop, EKWB will be at the top of my list. Bottom line, as frustrating as this has been, one faulty o-ring in one specific cpu block doesn't ruin the entire companies product line credibility. Mistakes happen, they stood up and admitted it and are trying to make it right. I respect that.

Not really. EK says I have to ship the unit back to them and they will pay the shipping cost, they said they will even arrange to have it picked up from my apartment, but when I try to get back to them to arrange a date/time for that, I don't hear anything back. Very frustrating. Right now the (defective) Predator 360 is just sitting in my closet back in the original package. As for the rest of the components: I still haven't 'tested' the motherboard. I just don't think it's worth the risk of potentially damaging the cpu/ram/gpu just to test if the board works. There is a big stain on the PCB of the board from the coolant leak. Yes I cleaned it off as best I could with 99% alcohol, but it's still stained. I can see that the copper contacts/traces that got wet look like theyve started to corrode or rust. MAYBE it will still work, but I've written it off as dead. Also, my 980 ti has a full cover water-block and backplate from EKWB, with the QDC connectors (it was in the loop with my Predator 360 before all this happened). I went to take off the EK stuff and just put the stock cooler back on, and one of the allen key screws on the backplate got stuck! So, if I can't get that backplate off, I can't take the water-block off, and I have no way of using that card if it isn't connected to a Predator unit with those QDC things. So essentially my 980ti is a paper-weight to me. I've got $900 into that card and I literally can't even use it. I'm debating trying to sell it on E-bay "as-is" and see if anyone wants it. The only way it can be run is to be put in a system with a Predator cooling unit. All in all this experience was really frustrating, and a very expensive failure. I've basically scrapped this whole x99 build and I'm just building a Skylake build instead, new motherboard, cpu, and gpu. Going back to just air-cooling for all my components. I did learn one valuable lesson though: open-loop water cooling is really cool, looks awesome, and the performance is really impressive (when it isn't leaking), BUT, I'm not going to invest in an open-loop liquid cooling set-up ever again unless I can afford to replace everything in that system at a moments notice if something goes wrong. tl;dr Basically this whole mess has turned me off from liquid cooling and I'm going back to air cooling. Lost a lot of money in this incident, still haven't got anywhere with EK customer support.

The original E-mail I got from EKWB stated that it only affected units manufactured after Nov. 1st. So as long as your MX block was made/assembled before then, I'm guessing it's unaffected by the o-ring issue.

I don't think any physical damage was done to the graphics card, like I said, the system never shorted out or anything and it was running WHILE the CPU block was leaking. I cleaned off as much of the spilled coolant as I could see from the graphics card using 99% alcohol, and currently letting it sit out to dry. The problem with the graphics card is that it has a water block on it, pre-filled with the QDC connectors, so I can't just throw it in another system and check to see if it's okay. I have to wait for a replacement Predator 360 to show up before I can find out if my graphics card that I've got $900 invested in is now a paper-weight... As for the motherboard, I'm still trying to get all the spilled coolant off of it, it seeped underneath where the onboard battery is, and I can't seem to get it all out, every time I try spraying it with compressed air, more and more coolant comes out. Honestly, as of right now I've all but written the motherboard off as dead. I don't feel comfortable putting my 5820k back into that socket and even trying to let it post. I mean, maybe it's okay, but if something does short-circuit, I don't want to lost my $400 CPU as well as my $225 motherboard.

So I got an email about potentially defective Predator units from EKWB. Shut down my system to investigate, and sure enough, the cpu block was leaking. But it gets better, it leaked out of the CPU block and down onto the graphics card pcb, right near where it plugs into the PCI-E lane, and got coolant INSIDE the PCI-E port! But it kept going, the coolant kept dripping down the motherboard PCB all the way down to the USB connectors, corroding every copper trace it came into contact with. I've totally dismantled the build and the motherboard is currently sitting on my workbench, but for the life of me, I can't figure out how to dry it out. What leaked out isn't just water, it's whatever coolant was inside the Predator AIO unit, and if im not mistaken these coolants are designed NOT to evaporate, correct? Honestly, even if I can get the motherboard 100% dry, I think the damage may already have been done as far as corroding the traces. As for the graphics card, it's a 980ti, that I have a full-cover EK block on, pre-filled with the Quick-disconnect couplings so it was in a loop with the predator 360. So even if the graphics card wasn't damaged at all by the liquid, it's basically a paperweight if it isn't connected to a Predator 360 with the quick-disconnect couplers. So my question is, anyone have any experience in dealing with EKWB customer service? The original email I got about the potentially defective unit said they will be giving free replacements for defective units, but can I expect them to do anything at all about the motherboard and/or GPU? This really sucks, this was my first experience with a (semi) custom water-cooling loop. I've heard nothing but good things about EKWB. I was so nervous putting the water-block on the 980ti, and installing the predator 360 was nerve-racking, and I was so nervous about screwing something up the whole time, now it turns out the damn thing leaks all over my system not because of some noob mistake I made, but because of a defective o-ring!

Thanks everyone for the replies, I guess since it doesn't really affect performance I'll just live with it, I just thought it was odd that a single gpu build wouldn't be running at full x16 speed.

Hello, hoping someone can offer some help/advice here. I just recently upgraded my mb/cpu/ram/OS drive and I'm having an issue now with my gpu not running at the full PCI bandwidth that it should. My hardware setup is: MSI x99A Krait Edition 5820k 32gb DDR4 980ti (single card) intel 750 series (add in card) and a sound blaster z audio card I first noticed the x8 issue in GPU-Z, and yes, I already tried running the render test inside GPU-Z to see if that would change it, didn't have any effect. If i go into the BIOS I can look at the motherboard-monitor screen and it shows me what PCI lanes are active, what's plugged into those lanes, and what speeds they are running at. The info I get from this is: 980ti running at x8 Intel 750 running at x4 Sound blaster Z running at x1 It's a 5820k... I'm not running SLI, I can't imagine why it wouldn't natively be running at the full x16 speed, I'm not even coming close to using all the available PCI lanes I have. Any theories? Thanks for your time reading this, I appreciate any help anyone can offer!

I can verify the ram issue. I just put a predator 360 in my Define S and I had to take the heatspreaders off my Corsair Vengeance high profile sticks in order to get it to fit.

Thanks everyone for the advice. Guess I'll just live with it as I've never flashed a bios on a video card before and I'd really hate to turn a 980ti into a paperweight.

Okay, I tried adding more voltage. If I add anything more than +15, Valley stutters and crashes within a minute. I tried +25 and it crashed almost immediately. Is this just limitations of the board/bios? I'll start looking into flashing it with a custome bios. Never messed with a video card bios before, and this card doesn't support dual bios.

I'm having some issues with overclocking and I'm hoping someone can help. I just recently got the new EKWB Predator 360, and a EKWB full cover water block (with quick disconnects) for my 980 ti. After leak checking and ensuring the new water cooling worked, I wanted to put that full cover water block to work and see how far I can push my 980 ti. Using this video as a guide: I followed Jays instructions and was hoping to get some decent overclocking results, but after several hours of working on this I'm sitting at the following overclock setting in Afterburner: Core voltage: +10mv (it wasn't stable in Valley unless I did this) Power Limit: 110 Core Clock: +100 Mem Clock: +150 Thats it. If I raise ANY of those settings beyond that, the overclock becomes unstable and crashes. My 980ti has a default core clock of 1140 with a boost of 1228, so with my 100 mhz overclock I'm able to get an actual MAX of 1425 mhz (measured with afterburner) while running either Valley or 3DMark. Hottest temp I ever saw was 40 degrees C on the GPU. The attached screenshots are of my current settings. *Note, my 980 ti is an EVGA hybrid model, so I obviously took the hybrid cooling unit off and put the full cover block on. I know the Hybrid card uses a reference PCB, so maybe that stock PCB is the weak link? Bottom line, this just seems like a REALLY weak overclock, I was hoping for a lot more than that given the awesome cooling from the full cover waterblock. Full system specs, if this helps: Asus Sabertooth z97 Mark 1 4670k, running at 4.0 ghz 16 gb ram 850 Evo 500gb Corsair Ax760i (using two cable for the video card, one for each connector) EKWB Predator 360 with GPU block in the loop. Win 10 Pro x64 Single monitor, Asus VG248QE running at 144hz Using the latest nvidia drivers If anyone sees anything glaringly obvious that I'm missing here, feel free to point it out. I'm very new to overclocking, so I'm hoping I'm just overlooking something simple. If I really did just get a weak GPU/card, that's fine, I just want to rule out anything I could be doing wrong. Thanks for any help, I appreciate anyone taking the time to read all that.