Dancop Posted July 1, 2015 Posted July 1, 2015 J518a728 4500/4500/3000c10 1.134v And this in the picture at watercooling! But I stopped it since it got too hot! [/img] Quote
rtsurfer Posted July 1, 2015 Posted July 1, 2015 Are we having a Devil's Canyon moment with 5960X's..? All J batches seem to be awesome on ambient, only seen 1 bad one till now. Here's mine J513B048 Quote
Paran0id Posted July 2, 2015 Posted July 2, 2015 Not a bad result, rtsurfer. Waiting for me to come J513B048, hopefully not worse. Quote
rsnubje Posted July 3, 2015 Posted July 3, 2015 (edited) Today I have this batch incoming: J518A728. Let's hope it does about the same as Dancop's Though I doubt it will do exactly the same. J batches seems to be decent. [EDIT] Just tested the 5960X (batch J518A728) on an MSI X99A Xpower, CM Hyper 612 and Scythe GT5400 @ 28c ambient. 4500@1.14v, uncore 4500@1.15v Edited July 3, 2015 by rsnubje Quote
rsnubje Posted July 8, 2015 Posted July 8, 2015 Currently I seem to be limited by voltage, at 1.6v. Is there a way to go past it? If I go above, it simply doesn't boot into windows, or when I adjust in windows, it will just crash. Quote
sabishiihito Posted July 12, 2015 Posted July 12, 2015 *Had* a J513B06 chip, did 4.6GHz 1.3v 100% stable but it somehow died during a memory bench session this morning. Nothing crazy either, just trying some Samsung-based Dominators at 3400C16. Board halts at Code 62 PCI initialize, swapped mobos just to be sure. Different CPU POSTs just fine. At least I'm within the 15 day return period! Quote
Guest george.kokovinis Posted July 12, 2015 Posted July 12, 2015 *Had* a J513B06 chip, did 4.6GHz 1.3v 100% stable but it somehow died during a memory bench session this morning. Nothing crazy either, just trying some Samsung-based Dominators at 3400C16. Board halts at Code 62 PCI initialize, swapped mobos just to be sure. Different CPU POSTs just fine. At least I'm within the 15 day return period! You are not alone. My 5960X died also three days ago. Died out of the blue under a normal reboot !!! It is the third 5960X dying in an 8 months time frame. I am lucky that Intel seems to love me ( ??? ) and keeps sending new ones for free. I am throwing away my RVE and I am going with Gigabyte. I strongly suspect that this board has severe issues. I have another problem now with this ...thing ( RVE ). It has decided to populate the four pcie vga slots as 16/1/8/8 instead of 16/8/8/8. So my one Titan X is suffering severely. Contacted Asus and wrote on their forums too but as USUAL they are either deaf or stupid asking me if EZ plug is in place !!! As for 5960X dying frequently. I have come to three conclusions : 1) Memory controller, deriving from Xenon family, is not strong enough in the long term. 2) Any sustained voltage use above 1.48 on mems deteriorates exponentially fast the life of the chip 3) VccSA above 1.15 is also dangerous. In all my three dead chips, as Intel advised after inspecting them, it was the memory controller that died. And this on 13/14/16/16/1T/270 trfc and 1.49V at 3200 G.Skill. With 4.6 uncore and 1.45V. COME ON INTEL !!! Maybe that is the reason they keep sending new ones with out much hassle. But this will end soon and then I am not willing to keep paying 1200 Euros every two months for the hell of it. I prefer to quit benching. My ultra expensive sports cars hobby costs me less. Lol. Quote
aerotracks Posted July 12, 2015 Posted July 12, 2015 You are not alone.As for 5960X dying frequently. I have come to three conclusions : 1) Memory controller, deriving from Xenon family, is not strong enough in the long term. 2) Any sustained voltage use above 1.48 on mems deteriorates exponentially fast the life of the chip 3) VccSA above 1.15 is also dangerous. Sorry to hear about your chip. Regarding my (not very well clocking) 5960X, yes to all the above, unfortunately, it still runs like on day one. Don't get me wrong, I'm not trying to kill it deliberately - but so far no visible degradation benching at up to 1.9V VDIMM (1.55V for 24/7) and 1.2-1.3V VCCSA. I have never touched cache though since I had a SLI Plus and 3500 was the end of it P.S. Xeon Quote
sabishiihito Posted July 12, 2015 Posted July 12, 2015 You are not alone. My 5960X died also three days ago. Died out of the blue under a normal reboot !!! It is the third 5960X dying in an 8 months time frame. I am lucky that Intel seems to love me ( ??? ) and keeps sending new ones for free. I am throwing away my RVE and I am going with Gigabyte. I strongly suspect that this board has severe issues. I have another problem now with this ...thing ( RVE ). It has decided to populate the four pcie vga slots as 16/1/8/8 instead of 16/8/8/8. So my one Titan X is suffering severely. Contacted Asus and wrote on their forums too but as USUAL they are either deaf or stupid asking me if EZ plug is in place !!! As for 5960X dying frequently. I have come to three conclusions : 1) Memory controller, deriving from Xenon family, is not strong enough in the long term. 2) Any sustained voltage use above 1.48 on mems deteriorates exponentially fast the life of the chip 3) VccSA above 1.15 is also dangerous. In all my three dead chips, as Intel advised after inspecting them, it was the memory controller that died. And this on 13/14/16/16/1T/270 trfc and 1.49V at 3200 G.Skill. With 4.6 uncore and 1.45V. COME ON INTEL !!! Maybe that is the reason they keep sending new ones with out much hassle. But this will end soon and then I am not willing to keep paying 1200 Euros every two months for the hell of it. I prefer to quit benching. My ultra expensive sports cars hobby costs me less. Lol. Very interesting, my chip was running on a R5E as well when it died. I also had a 5930k's IMC die on an X99 Deluxe a while back: http://forum.hwbot.org/showthread.php?p=345081&highlight=died#post345081 I may just have to lock VCCSA voltage down to 1.15 anytime I'm benching. I'm considering ditching Asus X99 too, I have an MSI X99A Xpower I can use. Quote
Guest george.kokovinis Posted July 12, 2015 Posted July 12, 2015 Sorry to hear about your chip. Regarding my (not very well clocking) 5960X, yes to all the above, unfortunately, it still runs like on day one.Don't get me wrong, I'm not trying to kill it deliberately - but so far no visible degradation benching at up to 1.9V VDIMM (1.55V for 24/7) and 1.2-1.3V VCCSA. I have never touched cache though since I had a SLI Plus and 3500 was the end of it P.S. Xeon Yes, Xeon, typo error. Indeed the one and only board that has not killed any 5960X yet is my MSI Gaming 9 ACK. Maybe it is the combination of high uncore and mem timings and voltages that creates the issue. Maybe RVE is garbage after all... Maybe... But I want to reassure you that the report came directly from Intel-Europe. Three dead 5960X, all due to memory controller. They are not lying. At the moment I am rebuilding an MSI based rig ( for fun ) with a 5930K that I have here and I will install 4 gtx980 in it and see how it goes. Anyway, my new 5960X is arriving tomorrow and I keep my fingers crossed to be as good as the previous one. My Gigabyte soc champion is arriving Thursday. CPU will not go on RVE anymore. RVE is going to the garbage bin. I said the exact same thing to Chino and Akane at the Asus forums. Kiss my ass goodbye. Two RVE's in 8 months. First defective. Never replaced or warranty honored. Second populates pcie slot as she likes. Yikes. I emptied my garbage bin an hour ago. I think there is enough space there for an RVE box. People, do not get me wrong. I have basted my balls for three days now trying to exclude peripheral reasons that might prohibit 40 lanes from working, including socket pins. Nothing apparent or less apparent. Even with one card in slot 2, it populates as x1. Nice !!! Bye for now. I have an itch to ditch this thing in my garbage bin. I am going... Quote
aerotracks Posted July 12, 2015 Posted July 12, 2015 Yup, same here, my experience with the MSI board has been very pleasant. Will try a Rampage next, should anything happen a RMA could only improve my situtation. I can understand your frustration, I'd still eBay the board as defective though Good luck with your next chip! Quote
steponz Posted July 13, 2015 Posted July 13, 2015 Most records .. if not all are on gb champion.. why bench anything else. 62 is usually gpu also.. Sorry to hear all the cpu killing.. haven't killed 1 yet.. have only benched gb. 1 Quote
Crew Leeghoofd Posted July 13, 2015 Crew Posted July 13, 2015 (edited) benched all but Asrock didn't have any dying on me (yet)... Just for info, running auto voltages has never been a good idea.. also those that run daily setups on water I don't think it is neither a good idea to push 4500 uncore and co for 24/7... there's hardly any gain for daily apps above 3.7Ghz... and sorry 1.45 Cache voltage is your own choice... What color is that when you set it at ? Red, yep it is... Honestly do you drive your sports car 12 hours a day flat out ? And you expect it to keep it for years... right... Edited July 13, 2015 by Leeghoofd Quote
kimir Posted July 13, 2015 Posted July 13, 2015 J513B036 CB15 4.5Ghz with 1.15v CB15 4.7Ghz with 1.3v Started playing with the cache after that, 4.2Ghz@1.2 is ok so far. Quote
Guest george.kokovinis Posted July 13, 2015 Posted July 13, 2015 benched all but Asrock didn't have any dying on me (yet)... Just for info, running auto voltages has never been a good idea.. also those that run daily setups on water I don't think it is neither a good idea to push 4500 uncore and co for 24/7... there's hardly any gain for daily apps above 3.7Ghz... and sorry 1.45 Cache voltage is your own choice... What color is that when you set it at ? Red, yep it is... Honestly do you drive your sports car 12 hours a day flat out ? And you expect it to keep it for years... right... Obviously referring to me. Sarcasm is NOT a privilege nor a wise way of communication with others. So, I am obliged to return the compliment and answer likewise. 1) Did I say I am running auto voltages ? NO. 2) Did I say that these settings ( uncore 4.6 with 1.45V ) are for 24/7 ? NO 3) Is 1.45V for 4.6 uncore required for benching ( few minutes benching ) ? YES. 4) Do you know REALLY anything about Ferrari's, Lambo's and Buggati ? OBVIOUSLY NOT. I most certainly expect ANY 5960X to survive for at least six months with half an hour of benching PER WEEK under conditions totally acceptable for a very powerful water setup. That is all. Keep it civilized man, whoever you are. We are here for fun and not some contest "mine is bigger than yours" RIGHT ??? Quote
Alex@ro Posted July 13, 2015 Posted July 13, 2015 (edited) Honestly this appears a user case problem.Not many people in the world manage to kill 3 5960x without using extreme cooling lol. Have you ever figured that maybe you set something wrong?Maybe you tighten the screws very hard on wb,etc.Killing 3 5960X has 2 possible choiches : a defective motherboard which happens time to time and then you should send it to RMA .b) your own mistakes which results in no possible fix. Edited July 13, 2015 by Alex@ro Quote
Crew Leeghoofd Posted July 13, 2015 Crew Posted July 13, 2015 You can take my comment/advise in any way you want man. Just a small part served as a reply to some of your statements listed in your previous comments. Again I have tested and benched all boards/CPUs on LN2 besides ASRock. No matter the brand, you dial in the voltages and you have to take into consideration the risks involved. If you bench 4700 on 1.45 Cache voltage it can go down. In my personal opinion that voltage is already way out of whack as it indicates a weak CPU/IMC/... combo. The CPU needs better than water cooling to handle these voltages during hours of bench sessions. If you are able to perform benching in units of minutes, mmm okay... OCing is all about balancing stuff. If you had 2-3 casualties I would revise the way of approaching things and start asking myself some questions... Easiest route is to blame the hardware ofc... @Steponzi: Sorry Joe two CPUs went out on your fav board at the Asia gathering, moist issues, don't you love them It is and remains a technical hobby, stuff can go wrong especially when pushed hard I'm not high and mighty, I flush the same stuff down the toilet is anyone else... Just handing out tips.. if you feel personally attacked, I must have touched a nerve... I just hope everybody learns from their experiences... Quote
Crew Leeghoofd Posted July 13, 2015 Crew Posted July 13, 2015 (edited) K good to hear Bala, one down than Could we get back on topic and list batches and results plz. Will update the excel tonite Edited July 13, 2015 by Leeghoofd Quote
Guest george.kokovinis Posted July 13, 2015 Posted July 13, 2015 Honestly this appears a user case problem.Not many people in the world manage to kill 3 5960x without using extreme cooling lol. Have you ever figured that maybe you set something wrong?Maybe you tighten the screws very hard on wb,etc.Killing 3 5960X has 2 possible choiches : a defective motherboard which happens time to time and then you should send it to RMA .b) your own mistakes which results in no possible fix. Nice edit. Stupidity-mistakes. Ha-ha... I will take the stupidity as most possible case scenario. Let the thread continue with what it is meant to be... Quote
Xtreme Addict Posted July 13, 2015 Posted July 13, 2015 So far I haven't killed myself any Haswell-E and I was using Asus, Asrock, Gigabyte, MSI boards, and people tend to say I am "xtreme killing machine" It's possible that simply your board might be bunnyed up and kills CPUs, it happens on all vendors, just sample NOT A MODEL problem. Quote
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.