View Full Version : Ccb2e 0608 FX60 Overclocking Woes?
TransNone13
03-22-2006, 07:54 PM
Ok, Ive been at this for 5hrs. I can't get this stable at 2.8Ghz. Dual SP2004 that is.
I tried just about everything I know of. I'm sure most of you either know more, or I'm missing something. I've tried anywhere from 1.3 -> 1.5v. I have an Xp-90c and a panaflow, temps below 50C. I know that sucks, but im sorry.
:brick: :help:
I have a Venus and I'm running mem stock for now 200 x 14.
At this point all I want is 2.8 :(.
Absolute_0
03-22-2006, 08:01 PM
2.8 woulda been a lot easier with a 300$ 3800+ and a 500$ phase change unit.
Anyway.. you've tried everything? I don't want to state the obvious, but all the ram dividers, all the LDT multipliers, and of course try changing FSB instead of just raising the multiplier... try many HTT/multiplier combinations, try more volts again. Tell us how it scales, like what voltage do you need to do 2.75? And what kind of ram is it?
TransNone13
03-22-2006, 08:03 PM
I have OCZ PC4000 2GB EB Platinum.
Would upping the FSB possibly help over raising the multi?
I'm currently running 2813Mhz 1.424v going on 10 minutes lol. Oddly enough, these tests usually fail around 12 minutes.
Ok, Ive been at this for 5hrs. I can't get this stable at 2.8Ghz. Dual SP2004 that is.
I tried just about everything I know of. I'm sure most of you either know more, or I'm missing something. I've tried anywhere from 1.3 -> 1.5v. I have an Xp-90c and a panaflow, temps below 50C. I know that sucks, but im sorry.
:brick: :help:
I have a Venus and I'm running mem stock for now 200 x 14.
At this point all I want is 2.8 :(.
Temps are 50C??? OMG that IHS Contact must suck....
fhpchris
03-22-2006, 08:08 PM
With a IHS that bad, that chip would prolly work well under phase.
TransNone13
03-22-2006, 08:10 PM
Idk if its just that, this HS is getting waaarm.
How much voltage were you giving it when it was hitting 50C? and at what clock? Was it at 2.8?
TransNone13
03-22-2006, 08:15 PM
Hits 50C with 1.42v+.
Oh my...That is very terrible...I would suggest maybe removing IHS if you don't care about the Warranty...of if it was OEM then it doesn't matter...but that seems very poor. That is the first 06 chip I have heard of with poor contact...
TransNone13
03-22-2006, 08:23 PM
I reseated the thing atleast 5 times. Does this chip seem like a possible defunct chip? I've considered RMA. :-/
TransNone13
03-22-2006, 08:36 PM
I'm trying 255 x 11 and I'm working my way back up on the volts from there.
Nanometer
03-22-2006, 08:45 PM
In a closed case with those volts, 50c is about normal. The xp-90c is designed to have lots of air flow, and lots of fresh cool air flow. Leave your case open to see if it helps any.
Don't up the volts, they are fine. Up the multi on stock HTT and see how high it goes. Give the chipset about 1.57 volts, and loosen mem timmings. What is your RAM typ? Oh, and what your chipset and pwm temps.
You might want to run single channel just for now, just for kicks. You can add it back in latter.
Im taking you have a powerful PSU, if you aren't sure tell us what it is and your system config. Speaking of power, make sure you use all the possible power connections available on the board. It's easy to forget one.
Coroner Kyle
03-22-2006, 08:47 PM
You should be able to do this on stock volts. I could with my FX60, but I know all chips are different.
Start from stock volts and speed. Then go up a little in speed say to 202. Then another few MHz until you are failing Dual 32M SPi with stock volts. Keep this going until you start to fail and then just slightly bump volts like by like .1 to 1.36 and see if you fail if you do slightly bump volts again maybe by .05 this time. Make sure you keep your memory below it's rated 250 so that we can rule it out. After you get to 2800 dual spi 32m stable. Then try prime. Also are you running dual small fft's or what?
My FX60 responds well to very small bumps in volts and I would recommend you try to up the vcore as little as possible.
TransNone13
03-22-2006, 08:50 PM
Alright guys. I've been slowly bumping vcore up .02v per failure in SP2004 dual instances (stress cpu small ffts). I'm at 1.392v now trying for 2805Mhz.
Also I have an 850SSI and when should I try playing with chipset voltage? I'd think that if stock FSB and 14 multi couldn't do it, why would the chipset help?
TransNone13
03-22-2006, 08:51 PM
Btw, I really appreciate your guys' quick responses and helpful comment/nature. You guys are awesome.
Badge56
03-22-2006, 08:58 PM
What is your multi on HTT?
TransNone13
03-22-2006, 08:59 PM
It's set to 4.
Up to 1.408v. I'm not planning on it passing.
eBoy0
03-22-2006, 09:02 PM
It's set to 4.
Up to 1.408v. I'm not planning on it passing.
Try 3x
Absolute_0
03-22-2006, 09:03 PM
Oh my...That is very terrible...I would suggest maybe removing IHS if you don't care about the Warranty...of if it was OEM then it doesn't matter...but that seems very poor. That is the first 06 chip I have heard of with poor contact...
Keep in mind, dual cores are much hotter than single cores. Add to that the fact that the XP90 isn't the best cooler around.
@ the OP
Have you tried another HTT/multiplier combination yet? It happens often (so i am told) that certain CPUs don't like certain multipliers... just try a different combination. Run a ram divider and try default 13x220.
And yeha like he says ^^ always use 3x LDT multiplier
TransNone13
03-22-2006, 09:04 PM
Right now I'm giving 255 x 11 a shot. My HTT is 4.
About the XP-90C, any HS that are better are marginally by a tiny margin. I'm looking at going back to water son with an Apogee, D5, and dual BIPs or BIXs. But if I can't get 2.8 on air, it's probably not a good investment.
Absolute_0
03-22-2006, 09:05 PM
Please use a 3x multiplier, not 4x. Especially when going past 250 HTT. It's not like 3x is any slower, so just run 3x, always, that's what i do.
TransNone13
03-22-2006, 09:06 PM
Ok, I'll definately try that out if 4x doesn't fly.
Coroner Kyle
03-22-2006, 09:08 PM
I know that when I was playing with my FX60 on air and I was trying to get the max speed bootable into windows playing with the chipset volts did help, but again that was when I was going for the max speed. I heard some very experienced guys on here say to leave chipset volts alone and now I do most of the time, but you can bump it by I think about .1 and still be fine. FX60's are hot chips and on air they can be a real painl, unless you've got a small turbine blowing on it. ;)
Just be patient and work your way up. That's what I'm doing as we speak with mine on water and this time around I'm running a lot less volts with a lot more speed than the first time I had it in my water rig.
Also you got an N or T wafer?
Coroner Kyle
03-22-2006, 09:10 PM
Just set LDT multi to auto and let your Venus take care of it for now. Find your stability and then try to OC that.
TransNone13
03-22-2006, 09:17 PM
Ok well, I screwed my windows clock up and windows wants me to activate. I don't want to since I'm going to reinstall once I get stable. So I'm going to reinstall XP, then go to bed. I have an exam tomorrow :(. Thanks for your help, but don't forget about me tomorrow :D.
Coroner Kyle
03-22-2006, 09:22 PM
I've reinstalled stupid XP like million times b/c of the activation thing. All I want to do Mr. Gates is tweak my rig until it's stable and then do a fresh install, but instead of having 30 days I get like 2 or 5 or if I'm lucky 12 and then I start over again.
Good luck with the exam and hit us up tomorrow we will slap that chip of yours into shape.
TransNone13
03-22-2006, 09:30 PM
Thanks, I hope we can lol.
zinff
03-23-2006, 05:16 AM
Is it stable in other apps besides SP2004/Prime? I cannot explain how or why but my CCB2E 0548 RPMW FX-60 will fail SP2004 or Prime95 within 10 minutes of start it but the rig is 100% stable. I can run dual 32m until i am blue in the face, any 3D torture loop, and have also run F@H for about a week now with NO work unit failures.
Perhaps there is a bug somewhere's in Prime type programs with the mem controller revision that is on the FX-60, just a guess here. I remember a similar problem i think when Winchester cores first came out, people were having a problem getting prime stable at STOCK speed and volts.
For 24/7 use i have these settings. Don't know if they will work for you but who knows.
CPU VID = Auto
Vid Special = 102.4%
Chipset = 1.52
LDT Mult = 4x
LDT volts = 1.4
Combined with running at 250x12 i get 3G for 24/7 use on water. Temps aren't all that great either. Idle is 30c and max load is 39c on an Apogee block. Quite a bit warmer than my 2.8G 0550 UPMW even when the FX is using less volts.
If i increase the voltage ANY it will fail even though the temps stay the same. Guess thats the limit until i muster up the desire to lop the IHS off to get better temps.
stone_cold_Jimi
03-23-2006, 06:25 AM
2.8 woulda been a lot easier with a 300$ 3800+ and a 500$ phase change unit.
Just the 3800+ would do, in many cases...
When I get into crap like this, I use lots of auto everywhere and see what the system thinks it should be doing.
General rule... the HTT you don't want it to exceed 1000...
200FSB use 5
201-250 use 4
251-300 use 3...
and so forth.. I think I use 3 all the time no matter what and have had no problems...
Also make sure your RAM is not your problem either...lately when pushing high FSB with 1:1 that is what I have been having to tweak to pass Prime...just something you might want to consider...
Also consider...a Thermaltake Big Typhoon...those XP-90c aren't the best in the world... I won't use anything other than a BT.
TransNone13
03-23-2006, 06:41 AM
Next step is water, not another HS.
I too think it's odd that it always fails about the same time.
I'm going to see if it's stable using other means.
Using Super Pi Dual 32M. Then I'll install COD2 and DOW and see if those work.
If it is failing the same time everytime no matter what...I think it is your RAM...in my past experience that is what it was with me..I had an Opteron 148 that I know would do 3.0 @ only v1.475 but would fail everything even at 2.5...and it was due to my RAM and the chip not liking it at all....
TransNone13
03-23-2006, 07:21 AM
Hmm, well I'm Dual Pi 32M with a Sandra Arithmic Burn in.
SO far so good after 25 minutes.
TransNone13
03-23-2006, 07:40 AM
Oh many 32M Pi runs do you think are neccessary?
neoman
03-23-2006, 08:12 AM
TransNone13 it happened with me also :P Dont know what caused it
TransNone13
03-23-2006, 08:16 AM
Thats odd. I'm feeling a little better.
Although, my ram is weird. It doesn't like me to change the tras or trc from 7 clocks.
Running 255 1:1 3-3-7-2
zinff
03-23-2006, 08:23 AM
I ran 4 runs of 32m back to back, no errors so i called it quits. Haven't looked back since.
TransNone13
03-23-2006, 08:25 AM
I ran 4 runs of 32m back to back, no errors so i called it quits. Haven't looked back since.
Sounds like a plan.
I'm kinda in the same pos.. 0536 tho. can't get it stable over 2.86GHz, and I had mem at 1:3 divider and multi at 3 when trying. Running multi @ 5 now tho without a hickup.
Read about it in the sig-link, I have a few pics etc..
TransNone13
03-23-2006, 08:35 AM
Wow, your rig is nearly identical to mine lol.
Levish
03-23-2006, 08:42 AM
It takes me setting ram to 166mhz vcore to 1.45 specified in bios and 200x14 to run 2800mhz stable
Temps suck even with a nekkid fx-60 and a zalman 7000a, it's a hot chip. I've checked mount and interference and it just happens to fit without anything touching (barely).
TransNone13
03-23-2006, 10:08 AM
Finished quad 32M Pi runs. What do you think?
I'm kinda in the same pos.. 0536 tho. can't get it stable over 2.86GHz, and I had mem at 1:3 divider and multi at 3 when trying. Running multi @ 5 now tho without a hickup.
Read about it in the sig-link, I have a few pics etc..Its a good chip once you put it under phase. Heres 0536UPMW.....
..Evap. temp. sweet spot is -33C...
http://img29.imageshack.us/img29/9826/screenshot1499gb.jpg
@TransNone....Thats awesome chip you got there:toast:
mikenam
03-23-2006, 10:23 AM
how many volts? 1.3? not bad. but i expect more for an fx.
Usually expect 3.0 on air from an FX right?
TransNone13
03-23-2006, 03:33 PM
It's not stable at 2.8 even though it did quad 32M Super Pis. Any thoughts?
EmineM
03-24-2006, 10:53 AM
Is it a 0608 APMW?
Levish
03-24-2006, 10:59 AM
i need about 1.42-1.45 vcore to be stable at 2.8ghz with air
TransNone13
03-24-2006, 11:06 AM
Well i restarted my efforts and Im at 1.472/1.45 ish.
Well i restarted my efforts and Im at 1.472/1.45 ish.
If you check the link in the sig of mine, you can see that I'm at 2860mhz, 1.424V or something.. Mem @ 260Mhz, it's rock solid..
Your fx60 seems to be able to go to 2800 at lower voltage than me.. but I haven't tried ldt multi @ 3 yet, still running it at x5.. I had to have the voltage where it is now, to be able to get a stabil 2.8ghz.
Vcore set at 1.45 in bios.. so my expert+bios is overvolting with around 0.26, ie showing 0.26 more than it really is, if I am to belive that smartguardian and cpu-z shows the correct one, and doing the same with around 0.2 on RAM aswell...
TransNone13
03-24-2006, 12:36 PM
Im pretty confident it wont hit 2.8 unless burn in is real. For whatever reason 2750 seems my limit ATM. 1.396v
Absolute_0
03-24-2006, 12:39 PM
Can't hit 2.8 on an FX60? :eek: sell it quick
Levish
03-24-2006, 12:44 PM
I was able to run 2x prime95s for 10+ hours at 2.8 with 1.45vcore and the stock heatsink.
I would have went further and with more vcore but temps were getting a little beyond 55c.
Most of the 3ghz+ screenies of fx60's were with over 1.5vcore, probably more like 1.55 specified in bios with 1.5 effective.
TransNone13
03-24-2006, 02:38 PM
Wow, thats lame. I've gotten an RMA. Hopefully they'll give me a better one. I know its :banana::banana::banana::banana:ty, but I need to do what i need to do.
Indyxc
03-24-2006, 04:46 PM
Is it stable in other apps besides SP2004/Prime? I cannot explain how or why but my CCB2E 0548 RPMW FX-60 will fail SP2004 or Prime95 within 10 minutes of start it but the rig is 100% stable. I can run dual 32m until i am blue in the face, any 3D torture loop, and have also run F@H for about a week now with NO work unit failures.
Perhaps there is a bug somewhere's in Prime type programs with the mem controller revision that is on the FX-60, just a guess here. I remember a similar problem i think when Winchester cores first came out, people were having a problem getting prime stable at STOCK speed and volts.
For 24/7 use i have these settings. Don't know if they will work for you but who knows.
CPU VID = Auto
Vid Special = 102.4%
Chipset = 1.52
LDT Mult = 4x
LDT volts = 1.4
Combined with running at 250x12 i get 3G for 24/7 use on water. Temps aren't all that great either. Idle is 30c and max load is 39c on an Apogee block. Quite a bit warmer than my 2.8G 0550 UPMW even when the FX is using less volts.
If i increase the voltage ANY it will fail even though the temps stay the same. Guess thats the limit until i muster up the desire to lop the IHS off to get better temps.
I had the same problem with an FX-60 I owned. Dual Prime stable 2900 at only 1.4v. 2950 at 1.42v, after that no matter how much voltage I threw at it, even 1.6v it wouldn't make it to 3000.
However, it would NEVER crash in anything else at 3000. Dual SuperPI 32m, at least 15 times, 3dmark06, gaming for hours. Drove me crazy. Sold it, and luckily found an Opteron 170 that does 2970 dual prime stable.
My theory comes down to motherboards not properly supplying the higher wattage required by the FX-60 at 2900 mhz +. I noticed everyone that does well with the fX-60 has a later revision of Ultra-D, or a DFI-Expert. Likewise, most people get to 2900, but after that point dual prime stable is no go, while everything else is ok.
TransNone13
03-24-2006, 05:05 PM
I have a venus, so it shouldnt be that.
eva2000
03-24-2006, 06:17 PM
Next step is water, not another HS.
I too think it's odd that it always fails about the same time.
I'm going to see if it's stable using other means.
Using Super Pi Dual 32M. Then I'll install COD2 and DOW and see if those work.
definitely try water!
FX-60 CCB2E 0536RPMW here cooled with 1/4" OD tubed Hydrocool 200EX unit pics at http://www.fileshosts.com/watercooling/HC200/DFI_NF4_SLI_DR/html/ and DFI NF4 UR SLI-DR with 1KW PCP&C psu
have you tried the new beta version of Stressprime 2004 07/03/06 build ?
new version of stressprime 2004 (07/03/06) based on prime95 24.14 http://www.fileshosts.com/stressprime2004/sp2004exe_070306.zip
Idle & Blend with 875MB memory allocated to each instance of stressprime
http://www.fileshosts.com/DFI/NF4_SLI_D/results/FX60/HC200EX/OCZ/PC4000EB/2_1/1114_3/LDT4x/12x/250-250-3325-7-14-2223_1.48-1.3-1.5-2.64_1552_ds8dds2_8F5/stressprime_idle_tn.jpg (http://www.fileshosts.com/DFI/NF4_SLI_D/results/FX60/HC200EX/OCZ/PC4000EB/2_1/1114_3/LDT4x/12x/250-250-3325-7-14-2223_1.48-1.3-1.5-2.64_1552_ds8dds2_8F5/stressprime_idle.jpg) http://www.fileshosts.com/DFI/NF4_SLI_D/results/FX60/HC200EX/OCZ/PC4000EB/2_1/1114_3/LDT4x/12x/250-250-3325-7-14-2223_1.48-1.3-1.5-2.64_1552_ds8dds2_8F5/stressprime_blend_custom_875MB_tn.jpg (http://www.fileshosts.com/DFI/NF4_SLI_D/results/FX60/HC200EX/OCZ/PC4000EB/2_1/1114_3/LDT4x/12x/250-250-3325-7-14-2223_1.48-1.3-1.5-2.64_1552_ds8dds2_8F5/stressprime_blend_custom_875MB.jpg)
Dual blend load running
http://www.fileshosts.com/DFI/NF4_SLI_D/results/FX60/HC200EX/OCZ/PC4000EB/2_1/1114_3/LDT4x/12x/250-250-3325-7-14-2223_1.48-1.3-1.5-2.64_1552_ds8dds2_8F5/stressprime_blend_1_tn.jpg (http://www.fileshosts.com/DFI/NF4_SLI_D/results/FX60/HC200EX/OCZ/PC4000EB/2_1/1114_3/LDT4x/12x/250-250-3325-7-14-2223_1.48-1.3-1.5-2.64_1552_ds8dds2_8F5/stressprime_blend_1.jpg)
http://www.fileshosts.com/DFI/NF4_SLI_D/results/FX60/HC200EX/OCZ/PC4000EB/2_1/1114_3/LDT4x/12x/250-250-3325-7-14-2223_1.48-1.3-1.5-2.64_1552_ds8dds2_8F5/stressprime_blend_24hr27min_tn.jpg (http://www.fileshosts.com/DFI/NF4_SLI_D/results/FX60/HC200EX/OCZ/PC4000EB/2_1/1114_3/LDT4x/12x/250-250-3325-7-14-2223_1.48-1.3-1.5-2.64_1552_ds8dds2_8F5/stressprime_blend_24hr27min.jpg)
fareastgq
03-24-2006, 09:57 PM
no no no no, don't rma that chip, slice it open and put it on water... that's an excellent phase/water chip.. You'll be sorry if you rma, you could be doing 3 gigs like me on water or 3.5 like dumo on phase... jeez, if you rma it, we might as well just trade chips... there's nothing wrong with your chip except it sounds like that tin cap is limiting your oc. Bad contact.
vBulletin® v3.7.0, Copyright ©2000-2008, Jelsoft Enterprises Ltd.