Mine worked fine, and i think angry_games used one as well in the OC database at dfi-street.Quote:
Originally Posted by Hybrid
Printable View
Mine worked fine, and i think angry_games used one as well in the OC database at dfi-street.Quote:
Originally Posted by Hybrid
Thanks! :toast:
ever heard of SNDS? intel processors are still a little bit more fragile than amd processorsQuote:
Originally Posted by calcal
Jort I can't believe this - there are things popping left and right now! I am sorry to hear you also took a beating. Glad you'll be able to RMA - 2 of my CPUs aren't RMA'able FX57 has IHS off - X2 4800+ had been dropped before and has marks...Quote:
Originally Posted by Jort
I was just thinking of doing another run with a third board and an Opteron - I think I am going to pass.
DFI told me that the new FX57 was suposedly still working along with one of the boards - kinda doubt it - but I shall see.
I have also killed one FX-57 with Grouper board. I was doing some memory testing at that moment. The system froze when I was changing the multiplier (from 12x 200MHz to 13x 200MHz). After this the cpu was completely dead.
This happened few months ago.
macci has also killed two FX-57 cpu´s with the same board.
The first one died under cascade cooling, and the second one has a partially broken memorycontroller. The cpu works and overclocks like it did before, but only 2x 256MB sticks can be used. If he tries to use 2x 512MB, either the system won´t boot, or it crashes all the time (if it even can reach to windows).
Same board and memory works fine with another cpu.
mikeguava and Jort [H]ard]OCP posted an article on thier problems with the
DFI crossfire, here is a quoteCould this be the problem facing dead Cpu's?Quote:
Originally Posted by HardOCP
The whole article
Quote:
Originally Posted by The Stilt
Sorry Stilt - sounds like an epidemic right now with any new board out there!
Read this - makes sense for Macci's porblem with the second CPU.
http://www.xtremesystems.org/forums/...t=79509&page=1
1st FX57 and X2 died running Geil One BH @ ~ 3.4V or more...
Quote:
Originally Posted by Jort
Can you give us some more details on what memory etc. you were running? Did you contact DFI about this?
http://www.dfi-street.com/forum/showthread.php?t=27644
its all here.
before i am going to boot the mobo i'll contact dfi :)
All of you guys are killing CPUs with high VDIMM on the Grouper/Halibuts right? I know Macci was using BH5 or VX. I know Sampsa was using Mushkin Redline at 250 MHz, so it must have been at high VDIMM. I also know Sampsa was at default (low) Vcore. How many of you have been running with elevated Vcore when VDIMM is high? Also, isn't it strange that all the 90nm CPUs seem to be getting damaged?
Maybe you guys are taking high VDIMM for granted since it was somewhat fine with Rev CG. But remember rev CGs came with higher default core voltage AND more importantly was on a different process node.
Before everyone starts blaming the motherboards, realize what you are doing. If a CPU dies at default voltage (ALL of them, not just Vcore) then there is definitely a problem. If you are killing CPUs at elevated VDIMM, especially over 3V then it's a completely different story.
We have gotten failure analysis results of some of the CPUs that have died on us under the cascade (running high VDIMM and high Vcore) and they all had dead memory controllers. What does this tell me? If you under the cascade and/or run with high VDIMM (even with elevated Vcore) you might be asking for trouble. We have had CPUs run for many weeks under cascade cooling and high VDIMM and Vcore, others didn't last more than a day or two.
Quote:
Originally Posted by The Stilt
Grayskull- please don't make it sound that this is a known fact - now that Tony gave the info about what migh be going on. All of us have been using these kinds of voltage differentials on NF4 boards for quiet some time - and in my case on several test rigs in 24/7 setups.
You are right it seems to be targeting the 90s - something that should have been addressed in Beta testing - no?
If it was known to the board manufacturers that the voltage differntial is a known killer - why was there not a warning etc. out there?
What Tony has stated shouldn't come as a surprise (it doesn't surprise me at all :) ). It is a known fact (or should be) that whenever you overvoltage above the stated specifications, there are no guarantees. How can you expect motherboard vendors to warn you if there is no quantifiable amount of differential that is known to cause damage? As far as AMD is concerned, overvoltaging and overclocking is illegal and not recommended. Unless a specification is posted from AMD, anyone trying to guess at a number is doing just that, guessing. The real value depends upon internal ASIC design and can vary with process. ;) So we've gone from 130 nm processors, to 90 nm and soon they'll probably be going to a finer technology and things might change again.
Also, don't think that any vendors are extensively qualifying their boards across multiple processors at anything other than specified voltages. It's not happening as far as I know.
Last, just because a certain lot of processors might be fine, doesn't mean others will. Process variation as well as the undoubted process tweaks that are constantly happening to boost speed and yield could also affect things.
Bottom line, overvoltaging is risky and YMMV.
Quote:
Originally Posted by mikeguava
Here's some stock settings trouble with my board, if I change any setting at all in the bios it wont boot, also, if I try to run 2 dimms it won't boot using slots 2/4 and using ballistix. This is all at stock. My board is going back I guess I will try the Grouper board now. Kinda silly that I can't even turn off the intro x200 screen and have the bios work This is using the new bios. I'm gonna try a diff ati board now. Tired of Nvidia so I won't be getting one of those.
grayskull, you KNOW it's related to the boards. How many people lost chips like this on the ultra-D boards? Not many if any I can think of... I and MANY others ran high voltages (I personally have run 3.8vdimm and stock vcore and 3.8vdimm with 2vcore through 3 different 90nm CPUs on my ultra-D with no issues even with watercooling and with dry ice) without the things dying. CPUs are dropping like flys now.
If this was an issue with the chips and voltage settings we all should have lost hundreds of chips since the 90nm chips were released by AMD. Unless you get some engineers on here who know everything and anything about these boards and chips you're going to have a hell of a time trying to convince us overclockers who have done the exact same things with the same chips in older boards with no problems that this is a voltage issue. This is worst occurance of deaths I've seen since the A64s first came out and modded shuttle boards were consuming clawhammers.
Someone arrest Fugger, Shamino, Macci OPP, OPB & CO immediately!!!!Quote:
Originally Posted by Grayskull
All kidding aside - I can totally understand form where you are coming from.
But what if the user is setting all the recommended voltages for his combination of hardware - it quickly becomes a grey area.
In my case I combine a DFI RDX board, a AMD FX57 CPU and Geil One BH memory and set within all recommended operating voltages. well with the Geil Mem it is between 3.4V to 3.7V ( and fully covered under warranty). AMD says 1.4-1.45.
Now who is at fault ? The Memory's ? - The CPU? The User? or the board manufacturer who has to combine everthing in one pot?
Another good example - same thing for me in the beginning - I get a brand spanking new RDX board - I throw in my Geil One -the board will not post - nothing...I throw in TCCD - now I can post - wellonly after raising the votlage significantly I can post with my BH memory - and I heard that the board was optimized for UTT/BH - hmm.Quote:
Originally Posted by fareastgq
One more thing - I think we can be pretty sure at least from this wave of potentially completely unrelated (or realted ) occurances of 90mns deaths, it doesn't appear to have anything to do with the ATI chipset ( nor NF4 chipset) in itself.
I have had nothing but trouble with this board from the moment I got it :mad:
I am experiencing the exact same problems as this.
http://www.hardocp.com/article.html?art=ODg4
Quote:
Running Into Trouble
I took delivery of our DFI CrossFire motherboard and prepped it for testing. We almost immediately ran into problems. The first thing I noticed was that I could not save my BIOS settings, exit the BIOS, and then reboot. After attempting to save and leave the BIOS, the motherboard would freeze at a blank screen. Resetting the board would allow a reboot, but that did not always work; sometimes a hard power down was needed. The reboot issues did not stop there. Even once we got an OS installed, the motherboard still exhibited reboot issues where it would hang on POST. Sometimes it would hang at IDE detection, sometimes at Silicon Image RAID controller detection, and sometimes it would hang after setting the interrupt channels. Then again, sometimes it would reboot just fine, but most times it would not. Checking the DFI support forums, I found that I was not alone in experiencing these issues.
Sheik,
Yes, this appears to be a wide spread problem. My finger is exhausted from pushing the reset-switch all the time. WE NEED A NEW BIOS.
-k0nsl
Quote:
Originally Posted by Grayskull
my cpu died WITHOUT oc'ing, i coudn't even welcome the POST screen.
maybe its related to there "super" 4 phase power design.
anyone taken the time to look at the burned pins and compare them to the AMD processor sheets to see which pins exactly seem to be getting too much current?
If we can find that out we can tell which part of the board may be operating incorrectly (over/undervolting and/or pulling high/low for increased/decreased amperage).
I've just checked the pins against the s939 pinout. 2 of the pins appear to be Vss (ground), and one of the pins is Vddio (the DDR SDRAM I/O ring power supply, according to the datasheet :confused: )Quote:
Originally Posted by STEvil
oooo does sound like memory controllers.
Board # 3 running perfectly fine for 8 hours now.
I really hope, u won't burn any more CPU's with the new board mike.
I'm kind of excited what's gonna hapen with my board. Just waiting for X1800 XT to get all together.