b-w-d.net

Hi, regarding your CPU its worth trying to contact AMD. AMD (europe atleast) have been known in the past to replace A64 processors with up to 5 broken pins, I have had a good experience with this. I only had 1 broken pin but "5 broken pins" was something I was told by the customer support staff over the phone. Otherwise, might aswell send it off for repair with that other guy.. but an AMD RMA would be better since you then get warranty again. And yes yes.. people say manufacturers wont take back hardware thats damaged and its in breach of AMDs warranty.. but they do it. Infact the website RMA form includes an "appearance" drop down list that includes things like "burnt" "cracked" "burnt/cracked/broken pins", and selecting broken pins does NOT outright result in a failed RMA, they still took it in & sent a replacement a few weeks later (it was rather slow). And why would they do that? Because theres 939 pins to play with hehe Good luck with your kit

I know this is an old thread, but on reading that I had to pose this. I have also destroyed an X2 4400+, a pin got bent back on itself during installation & when I tried to bend it back it snapped off. My advice to you is, if its not too much hastle, its still worth trying to RMA. Despite the obvious physical damage involved in my case (although more minimal) AMD were very helpful and did replace the processor. Talking to the AMD customer support staff here in europe I was informed that AMD generally accept RMAs for CPUs with up to 5 broken pins (I even managed to get a reference on that, because my local supplier were being more than a little difficult). Hope you enjoy it anyway paperweight or otherwise

Incase anyone else is looking at this thread, I talked to Angry and am going to RMA the board. Hopefully we can consider this "resolved"

Small update from a vapochill user. Be careful with your chip insulation! Ive heard a lot of things about pin-rot, but ive seen a friend of mine taking apart one of his vapochill setups and one of the pins literally fell off, no other way to look at it. The freeze/thaw process loosened the pin to such an extent it fell away from the bottom of the processor (not a snap! its clean off right at the base!) To be honest, I see no way to protect against that, so be especially careful how you look after the processor. Having seen this.. im moving back to air.

To clarify, the point Lunchbox is making is this: Running more than one stick of ram in an A64 system naturally puts more stress on the memory controller in your processor. These memory controllers are rated for DDR400 (and no more). It is therefore POSSIBLE that while your sticks run fine individually at DDR580, your CPUs memory controller is incapable of handling the additional stress caused by running two of these sticks simultaneously at such high clocks. Ofcourse this is a quite defeatist and Id recommend you continue tweaking, hopefully everything will work out for the best. But it is possible that your CPU just cant handle two sticks at that speed. Note: this is the same reason its a bad idea to use all 4 slots. Running 4 sticks of ram in an A64 system rarely works out that well, usual outcome being a serious drop in bandwidth (running at 2T).

Exactly, the biggest issue I had with nF2 was bios corruption, beyond that it pretty much just worked. The transition to nF4 has been.. painful, whether its due to PSU connectivity, PSU power, complex memory settings, nvidia drivers etc..etc..etc.. To be fair though, this is the price we pay for performance. But yes, I too have spent a LOT of money on parts & had a lot of problems.

I have this exact same problem, and have also tested it in an SLi-DR motherboard with the same effect. No dual channel, ive tried optimised defaults and the timings shown in this thread

by "real life" i mean the voltage my RAM is reporting via its onboard circuitry and inbuilt LCD display. in the past this has given pretty accurate readings.. a discrepancy of .2v for vDIMM is a lot

with only one stick installed i can get up to 2.55v :/ .. but it still wont go higher

Hi I recently upgraded to an Expert, having previously used an SLi-DR, and prior to that the nF2 boards ("back in the day"). Ive been replacing a lot of my hardware and after installing the new motherboard i started it up and vDimm was ~2.5, but since I was upgrading the PSU next day I didnt think much of it. Today, swapped PSUs, and ofcourse the vDimm is still 2.5v, but the thing is changing the setting in bios seems to have minimal effect. Before anyone claims this is due to inaccuracies in the motherboards voltage readings im taking this value directly from the sensor in my ram (which reported voltages accurately in my SLi-DR setup). Essentially, setting 2.7v in bios = ~ 2.4-2.5v real life setting 2.45v in bios = ~ 2.39-2.45v real life It seems impossible to get over 2.5v (i dont want to push the settings too far incase it does actually set the voltage correctly and blows the ram), and considering one of the Experts features included the removal of the 4V jumper (improvements to vDimm?).. something is aweful strange about this :/ All voltage rails are stable, with very minor fluctuation on 5V, I have 24pin ATX, 8pin 12V and the fdd power connector plugged in. What gives? I know corsair memory doesnt run nicely on DFIs motherboards since the move to 64bit, but running at 2.5v is just ludicrous, any attempt at serious processing results in BSOD. AFAIC once I get this dimm voltage increased i can sort out any remaining instability myself. I have tried the big-toe modded bios, and currently have the latest official release flashed - no improvements seen. The the ram is currently in slots 2 and 4 , think 1 and 3 wouldnt boot but id have to try again to be certain. any ideas?

I have had this exact same problem! No idea how to fix it

Thanks loads for your posts guys . haha, one of my mates is telling me to use a key-activated switch for it or maybe one of those covered aircraft style ones what u recon?

I wonder when someone will come up with a decent 3rd party kit that pulls the air out through the PCI brackets below the cards! Id use them with my 7800s in SLI!, if they designed it reasonably well there should be enough clearance for the airflow too. Hope Artic Cooling get on this quick!

Hi there With my SLI-DR, as im sure others are aware, the CMOS reset jumpers are hidden between the two graphics slots. This makes it very difficult to clear CMOS with a fully installed system. My plan is to make a switch for the jumpers so I can clear CMOS without having to use a pair of tweasers and a torch every time. Now could you someone (preferably of reasonably authority) please tell me, are the 3 pins absolutely necessary or is one of them only there to hold the jumper in place when NOT clearing CMOS? The reason I ask, is if only 2 of the pins are actually doing anything then I can use a normal 2-pin ATX switch on the 1st and 2nd pin for this, whereas if a connection is required between 2nd and 3rd pin for normal opperation of the board then I need to wire my own switch myself. Thanks

I know exactly what you mean. Noone gave the writer of the A-list the authority to dictate that any other hardware will not meet the necessary specifications. This PSU issue seems to come up in 100s of threads, most of which, by the looks of it, they arent relevant to. People should encourage diagnosis of a system, not wrecklessly throwing money away for new parts without checking if the old parts are even at fault! I am sick and tired of seeing this from forum users and moderators alike. This is counter-productive people! Stop cutting corners!