Geeee, hope your findings will avoid numerous cpu deaths, good job there.
Printable View
Geeee, hope your findings will avoid numerous cpu deaths, good job there.
Mike, another one dead???? Well, hopefully the 4600's still OK :p: Hey, BTW...you gotta PM comin ;)Quote:
Originally Posted by mikeguava
Quote:
Originally Posted by s7e9h3n
Nope - playing it save with crap CPUs for now - lol
That 4600+ is sitting in a safe with five locks - an tossed the keys in the ocean so that I am ot tempted to play with it...
Lol
LOL so that's the key i found on the shores of our Australian beach :DQuote:
Originally Posted by mikeguava
Quick Eva, send that key to me! Now if only I knew where in Miami Mike lived ;) .........Quote:
Originally Posted by eva2000
Got a new RDX200 here too, but I think its better stay in the box:(
I like my 55 claw 0519 "N' too much:) Its the last one..
Well I can kill CPUs now on demand!!! Pretty damn easy. ROFL. Just lost one...
I was very dissapointed when it became clear to me that DFI trouble investigation only went into "how dirty and abused" my CPUs were.
I had written them to look into specific things - but it became obvious that I had to do the work myself.
At the time during the investigation I was preoiccupied with far more serious troubles and didn't really take the time to be on DFI's balls, and my frustration with them only let me say things like " good job guys"
There is a mayor bug in the bios with the RDX - which is a certain cpu killer!
I just set a CPU to 1.25V + 104.8% on air on a brand new OPTY - and got 1.6-1.61V again. I'll have to look on where to upload the video - if anyone wants to donate another 939 - I could make a really fun video - something along the line of 1.0V + 133% -lol
The board which I had receive from DFI directly was supposedly tested for several days. It came with bios 10/11 - nothing had been changed or modded by me.
At first I had tried to take careful steps not to discredit DFI at all.
before the final result came back from DFI - I even made a silly post that I am already running another board which was a lie - - to reduce the already negative effect my post had created.
Me intitally posting really was intended as a warning - but now I am just pissed - because I feel that their "investigation" was solely a spin control on my back.
As far as the board goes - I will still use it - now I know how to avoid ( hopefully!!!) any further deaths - unless I decide to kill 100-200 X2 CPUs and send DFI the bill -just kidding
Guess it's time for Supreme Evil to look into things again
So 1.6-1.61V will kill an opty? Or, are you saying there other manual vcore combinations that you know of that are CPU killers? If so, what?
No - 1.6 won't kill an Opty - the Opty is still good.Quote:
Originally Posted by xenolith
It's not a specific combination of manually set Vcores ( this is what I had suspected at first - but now know better) it is ANY vcore plus the added % value after installing a new chip. This new boot of a new chip - the moment you set any manual value for Vcore will result no matter what 1.55V plus the added %. SO if you set something weird like 1.15 plus 113% - you'll end up pulling 1.7V - which on a X2 is a killer on air if you really put some load on the CPU
Thanks for letting us all know Mike.
DFI :nono:
1.7v kills? that seems a bit low?
Not on a low voltage X2 that runs fully loaded on AIR in 3DMarks where you're not monitoring the temps etc...both my 4600+ on 4800+ ( R.I.P.) do/did not like anything above 1.55V - even under extreme cold.Quote:
Originally Posted by STEvil
http://www.xtremesystems.org/forums/...ad.php?t=82185Quote:
Originally Posted by mikeguava
1,75v :|
Well, I guess it depends on the type of X2 - the stepping etc. If you have the low voltage kind of X2 ( some run 1.23V default - others run 1.3xV default ) it changes quickly. There's some magic weeks and steppings of X2 that behave differently than what is described in that thread and can under no circumstance run at those voltages. Especailly under Miami air..Quote:
Originally Posted by biohead
Quote:
Originally Posted by Dumo
Just read for the first time that you also lost a CPU on a RDX - care to share what happened in your case?
Heres Mike...http://www.xtremesystems.org/forums/...ighlight=dumosQuote:
Originally Posted by mikeguava
This after RDX and FX55claw stopped @ middle of Superpi benching with 1.8Vcore and 3.6Vdimm and Zippy 700w psu. After that it gave 4 leds on for a second and off...everytime I powered it.
Then why not lower the startup value in the bios?Quote:
The way this happens I am pretty sure is when you switch CPUs - the bios gets stuck in the start-up value ( 1.55V) and adds whatever % you have added. X2 4800 RIP! I will have to confirm that later - to find a sure way to reproduce this problem.
Tmod
Quote:
Originally Posted by Tmod
I haven't checked if the startup value will have any impact - but for some reason I have a feeling it would also not have any effect until the second reboot from manually setting it - just like the CPU VID CONTROL setting does not work until the 2nd reboot.
The best thing will be to avoid any VID SPECIAL CONTROL settings leave them at auto ( or 102.4%) and things should be safe. The worst would be 1.55V Vcore.
Quote:
Originally Posted by Dumo
Completely missed that! Sorry bud!
Are you going to try disabling or changing the startup value and see what it does?Quote:
I haven't checked if the startup value will have any impact - but for some reason I have a feeling it would also not have any effect until the second reboot from manually setting it - just like the CPU VID CONTROL setting does not work until the 2nd reboot.
The best thing will be to avoid any VID SPECIAL CONTROL settings leave them at auto ( or 102.4%) and things should be safe. The worst would be 1.55V Vcore.
I see there are the following settings.
CPU VID Startup
CPU VID Control
CPU VID Special
Between those three there should be something that can be done to help with this problem.
Tmod
manually set the startup value to 1.35 for an X2, its meant to be raised for cold boot on uber-clocked chips, but it may be raising due to some weird non-detection issue.
Board is running 100% stable for 24 hours - overclocked and under load - with reboots all the time. As long as no CPU is swtich the board run really nice
^^^ Mike
Should I worry about my San Diego after posting with a Winchester (Never changed vcore or VID Special to my knowledge) Should I not boot it up, or just wait?
Quote:
Originally Posted by JasonDTM
No need to wait - as long as you don't touch the Special control you're fine. Once evrything is settled you should be good for any changes if needed
BTW - Vcore in Smartguardian is 100% accurate for me - checked with digital Mutlimeter, and mem is just a tad bit low - same as perisvore's findings
I am not 100% cerain about this yet - but it's the moment that you change the VID Control manually after installing a different CPU - not just the first boot - but the first time you set the voltage manually. Haven't had time to verify this
Alright, is this a bios bug, or is it a componet bug? After monday I will not have another cpu, so it'd be SOL for awhile, The winchester I have is a little champ... has had 1.9v on air into and hasnt died just yet... lol
Think bios - can't say for sure, but a different bios could fix that...
No troubles... FX57 + RDX-CF + DI
with R.A03 RDX-CF (bios 2005/10/11) . Vcore was 1.550 + 116%... And that FX57 is still alive :D We also played with one Opteron but that was a huge disappointment (it's also still alive)
http://dfics.dfi.com.tw/dfi_cs//RDX2...ue/Photo_2.jpg
http://dfics.dfi.com.tw/dfi_cs//RDX2...ue/Photo_3.jpg
there was wire between the pins? :confused:
and yeah, those pics look amazing! :D
nice cam! :D
Truly amazing pictures! Very dramaticQuote:
Originally Posted by saaya
Took me for ever to have this one analyzed...
The residue is from acetone and dielectric grease solution - the chip had previously done a dry ice run done and had some die-electric grease residue left. Before shipping everything out at one point I started cleaning the CPU with acetone but stopped in the middle of it. What looks to be a wire in fact is just a residue of acetone and dielectric grease unless the line in pic # 2 was some type of fibre.
DFI has left out also pictures of tobacco within the pins and other debris - as the chips had been tossed also on a couch and picked up even more random Debris. But that would have been to obvious that the chip hadn't run in that condition..
Again the chip wasn't used like that...the dry ice run was done way before the troubles started...
Looks like we got everything figured out. I killed the cpus by clearing the cmos by using the "clear cmos" jumper in conjunction with also removing the battery. Actually only the "clear cmos" jumper should have been used - and NEVER the battery...ok...
http://www.dfi-street.com/forum/showthread.php?t=33668
oddd people the cmos jumper and battery procedure was recommded for prior NF4 boards, why the change now.
so what happens when your CMOS battery dies on you?
:fact:Quote:
Originally Posted by STEvil
i dont think anyone would keep these boards long enough for the battery to die. If the battery does die, DFI will simply say "Don't let your battery die, it will take the CPU with it...since your bios will get corrupted and it will get confused and randomly boot up with random settings" j/k
I think the best is to these:
1. Avoid removing CMOS battery. Powering off the system and moving the CMOS jumper from normal to clear is enough to reset the CMOS.
2. Do not power on and off right away. There might be some live electric charge on the motherboard and power supply capacitor that can produce electic spike.
The battery had no effect on the VID core bug so far...had the bug from the moment I started with the the board from DFI - first boot with all auto was fine - after setting the vcore manually the first time reboot gave the Vcore jump - without me ever removeing the battery at that point...Unless DFi had removed the battery before they sent it to me.
Quote:
Originally Posted by mikeguava
My sarcasm in that post wasn't really clear - but it turns out that the battery really does have an impact! I initially thought this was BS - because I had the Voltage BUG with 2 boards -where i hadn't removed the battery.
One of the boards was BRAND NEW, the other one was a RMA board from DFI - which was supposedly tested for 2 weeks. Turns out the DFI engeneer must have also used the battery removal procedure :rolleyes: .
Thanks to Saaya I looked into this one more time and it turns out, that the battery removal with or without the "clear cmos" jumper - this bug gets triggered.
So any user with a brand new board should be careful with the VID Special control. And also later on - not remove the battery if possible. And when you do - again make sure to keep the VID special control at auto or lowest settings...
OOPS -just relaized that I am running @ 1.7V while I am writing this - need to reboot before I share the results!!!!!!!!
hmmm so removing the cmos battery resets the system in a way voltages might be too high and kill the cpu and maybe the boad?
but a bios reset cant cause this?
i didnt know that removing the cmos battery resets the system differently/more than reseting the cmos.
i hope this helps to prevent more people from running into problems with the board :)
whats up with the special vcore thing though?
or do you mean the startup vcore?
i was under the impression the battery jumper just disconnected the battery from the cmos to allow its memory to clear to default settings?
That seems the simplest way to design it on a motherboard to me, anyways... no hassle, works every time....
Not in the case with the RDX
apparently.
I was looking into getting an Asus A8R-MVP, but now i'm wondering if this issue could span the entire range of ATI RDX chipset based boards (I sure as hell hope not, i've been waiting a long time for a new system!).
:(
I tried it on the Expert - but didn't see it in the one quick trial run. But I am too worried right now killing anything else with these boards- especially since DFI uses this Battery Removal to deny any claims, on what seems to be the industry standard to clear your cmos!!!
I think even DFI.....now confused:(
LOL.... i'm playing with RDX200 and my Opty 144 here and after reading this thread, it makes me feel scary.....
No big problem till now but the system sometime hang in bios and crash to black screen in Windows without turning back on. I have to power off the PSU then switch back to get it on again.....
My board is: D54016189
I have the same problems sometimes & have been using the board since the day thay came out, I was a bit warey at first but I haven't had any major problemsQuote:
Originally Posted by ValkyrieLenneth
There funny boards, but don't expect such high clocks on your mem:( !!!