Printable View
http://i787.photobucket.com/albums/y...KE/superpi.png
I'm going to run prime tonight while I sleep and if it's fine when I wake up I'm calling this stable. Linx has failed me even at bios failsafe defaults sometimes 1st run sometimes not until run 15, I will just have to live with not being in the 4.5 linx club. :p
I would try to drop the memory to 10x, set it to 9-9-9 1t and uncore to 20x, just to take them out of the equation. What is your qpi set to? Are all your v's set to what I was using? For benching you should be safe to go up to 1.4v for vcore and qpi/vtt. I run cpupll at 1.9v for benching. Keep at it, you will get there! :up:Quote:
Originally Posted by mbreslin
linx has failed at bios failsafe defaults steve, ram was at 1066 999
Are the Vref voltages suppose to be 1/2 the Vdimm?Quote:
Originally Posted by mbreslin
http://i787.photobucket.com/albums/y...OWPOKE/hot.png
could chipset heat be causing all this? I know I keep asking, because I don't want to drain loop until final fittings come in a couple days, when they do I will be able to reseat my mb block. Ever since first installing that 1 temp has always been super hot even at dead idle. I must have sloppily installed chipset block?
^^Yep, it could, and properly just did. 70'C idle is just way too much.
Use the silent pipe with a lite fan on it, and you'll get 52'C load, that's what I get, anyways.
Thanks sam, how do you think I could have messed up installing chipset block? Could too tight do this? Just trying to get some idea so I'll know how to proceed for reseating. Also, could you post a screenshot of speedfan showing your 'temp3' ? Thanks.
This is my Idle-temp, 47'C, with a hot GTX 295 in a closed case.
I have a fan on the silent pipe that keeps the chipset pretty much stable at 52'C under load.
http://img.techpowerup.org/100417/Capture035.jpg
I am still on stock intel cpu cooler for now, but i will probably go to a h50 from corsair.
I am loving my Windows 7 exerience on the C300 ssd.
Can't wait to get my cpu cooler and see what this cpu and gskill perfect storm kit can do at +4Ghz and DDR3-2000
again ... I am loving my ssd, but does anybody know WHY I get the following error from as ssd utility when trying to get the Acc.time Read :
Error! Could not open device \\.\PHYSICALDRIVE1
apart from that I am lovin.. oh yeah i said that already :p:
http://i105.photobucket.com/albums/m...sbench-web.jpg
I noticed you guys having a tough time saving your BIOSes. You can make a small 1GB or smaller FAT32 partition at the end of one of your hard drives if you want, then you can save to HDD from the F11 function in the BIOS or from Qflash.
You can also use @BIOS to save with, DO NOT USE TO FLASH WITH PLEASE!! This will save all profiles as well as current settings, so be sure if you share a BIOS and think the user might have issues with your current settings, be sure to tell them to load optimized after the flash before trying one of the saved profiles.
http://i41.tinypic.com/359a5mw.png
http://www.gigabyte.com.tw/Support/M...lity&FileID=52
@ Fritz
Easytune does have a uninstaller, it would be found in the start menu/gigabyte folder and also located in the installed program folder. Are you still having problems with this?
I have not got any of the new versions to work with UD7, have not tried the last 2 though I don't think. I have only been able to use a old version, which I uploaded for others who had the same issues as me with the newer ones (Freezing on apply).
Here is the version I use for UD7 (B9.1120.1 - (11/20/2009))
http://www.mediafire.com/?idi2vezyhcm
To the users mentioning failing LinX or other programs at Fail Safe, this is expected, you should not use Fail Safe for anything. Troubleshooting mayber, but even then I would suggest not. Use optimized Defaults to start with, then you will still need to set your own Ram and QPI/Vtt settings manually. Do not rely on XMP or Auto to get those correctly, there is no way they could guess what your CPU's IMC needs for any setting, you will need to find that on your own.
@ mbreslin
Are you using the EK FB UD7 block? If so then yes, it may not have any contact at all. I installed mine and found it to be very hot at idle with stock voltages and quickly shot up to over 75+ at stock with voltages manually set for testing. Removed the block and found ZERO Contact. I spoke with EK about it and they said I must have not had it on tight enough, I swear I almost broke it because it was so tight.
Anyway, no luck trying again either. I think something is wrong, like the spacers are just to large or the screws are too long causing them to not be able to fully screw in? Anyway, I swear it was going to break every time I put it on, the screws are WAY to hard to screw in and contact could never be made. I ended up putting on WAY to much paste, so I knew it would make contact, and then it finally did. So you may need to put on a huge excess of paste, that is the only way I could make it contact the chip.
Temps are now great, less than 36C most of the time at idle voltage/speed (Which I do not use), and 40-42C at 4.2Ghz with 1.2-1.24V set (if using 1.14-1.18V temps are 36-38C). I did not like the idea of using a huge amount of paste, but it seemed to be the only option other than trying without the spacers and risk cracking the chip, but it worked out fine in the end.
I think EK really needs to work on this, something is not right I am sure.
@Lsdmeasap: Thanks so much for your input, which do you think wasn't making contact? nb? sb? both? may be silly question but are those the only places you needed to put paste? Thanks.
I found the NB was not making any contact at all, even after a tried a few more times thinking it was going to break. So you are using the EK Full block then right?
I had to put way to much paste on for it to make contact, you know like the amount you would find used on stock GFX Cooling heatsinks (TOO MUCH). Then it was fine and made contact and temps are now good
Yes I am here is an older picture (using different cpu block now):
http://i787.photobucket.com/albums/y.../slowpoke4.jpg
I installed it and never really thought about it but I have noticed one temp (temp3 on speedfan, different names on other programs) has always been like 68+ at idle and 72+ or so at load. In a couple days I'm going to drain loop and pull it off and take a look, good to know I'm not the only one, and good to know it's fixable. I have plenty of mx-3 to use on it.
It is probably the NB. You can check to be sure by installing the latest Everest Beta, it shows the IOH/Northbridge Temps
This is 1.24V IOH.
http://i39.tinypic.com/96c6fk.png
You can get the latest Beta here
http://www.xtremesystems.org/forums/...=216918&page=7
Or a free 30 day trial of the latest final here (Not fully functional, so may not show NB temp, not sure?)
http://www.lavalys.com/support/downloads
Before when I first installed and noticed, if I used 1.24 the temp would shoot past 75C within a few minutes. I noticed as soon as I turned on the machine, at stock 1.1V it was 50C + and went to 70+ very quick if I did anything, I knew this was a problem because the stock heatpipe was only 46-48C always I think
I use TX-2 myself, but if you have MX-3 that is fine too. Costs so much though, I'd use MX-2 if you have any left since it's cheaper.
You're right, I can't wait to fix this now. This is after like 20 minutes of just idling:
http://i787.photobucket.com/albums/y...LOWPOKE/nb.png
Edit: Just wanted to add I'll be using mx-3 because I have plenty but also because the skinee labs tim roundup showed that mx3 actually gets better the more you use. I think by the end he used probably around what it must take to get contact with the block.
Ya, looks like you definitely have the same issue I had when I first installed it, NO Contact on the NB at all. You will see when you remove it, the block will still be clean under the NB
Not sure what the problem is, either spacers are a little too thick or the screws are not able to fully screw in causing no contact?
I'll do whatever it takes, I don't know where I put the stock thing anymore so I'm kind of screwed. It may have got thrown away. Hopefully the nb temp is what's causing linx to puke. Sometimes it fails even at bios defaults, cpu at stock and ram at 1066. I was really hoping my 980x would do 4.5ghz at ~1.4vcore for 24/7. It seems to do everything fine, prime for hours, superpi 1024m, memtest86 passes with all 6gb at 2000mhz 7-8-7-20. Linx seems to randomly fail.
It probably is, I'd be scared to run it on mine when I noticed. What were your NB temps under LinX load? What voltage are you using for the NB now?
http://i787.photobucket.com/albums/y...KE/nb_load.png
It doesn't get as high as you would think, I think it pukes before it can get very high. After an hour or 2 of prime95 it's still only 74ish. Still I'm sure that's way too hot for it. Hopefully I haven't done any serious damage from having it idle at 68, I asked about it in a post once, people said it was 'no big deal'.
edit: my post:the response:Quote:
Originally Posted by mbreslin
Quote:
Originally Posted by astrodanco
Well I think you may have better contact then I originally though, but still not the greatest it looks like. At 1.24-1.28V IOH using stock heatsink my temps did not get even close to that, after 5 passes LinX I think it was only 46-48 or so, not much more than idle temps of 40-42 and the 48 was at the higher voltages 1.28.
So I think you may have partial contact or it would be much hotter than it is now.
As for your quoted posts at the end, you have to keep in mind they may be thinking of older gen boards, that only had the "System or Motherboards" sensors, those were in the PCB not the IOH. These boards have both as you can see in Everest, one in the PCB near the NB, and one in the IOH/Northbridge itself
I think LinX may have some issues with 980X, aside from some i have seen mentioning core issues, I get random memory errors like Not enough memory when I choose 3GB tests (Also says this in the middle of a test sometimes). Sometimes on 1GB test it will tell me I do not have enough, when it is showing 4-5 available.
I think it needs some updating for 980X still, did you try IBT and get the same errors?
Ibt crashed on the second run. nb back up to 72+. Going to lay off until I get that sorted. Hope that's it.
Sorry, forgot to mention in my previous reply. I think the temps are still safe just not correct for what they should be with your block or even with stock cooling.
X58 is rated to 100C, so I don't think you would've done any damage, temps should be much better once you get proper contact.
Thanks for your reply!Quote:
Originally Posted by Lsdmeasap
What really amazes me is that I donīt have any problems on Vista to remove the old version and to install the new one in your link. As I do the very same on W 7 it wouldnīt work for some reason.
P.S. Sorry that I was kind of pi**** off in my last posting:cussing:
I just checked EK UD7 fullboard cooler and it didn't have good thermal contact with ICH10R, made spacer about fraction of mm thinner on all spacers and made full connection :up:Quote:
Originally Posted by Lsdmeasap