T_M

RAM

dinos22

true...he used 2x256MB which would be slower


you should buy one of those cheap Opteron 144 CPUs for some FX57 ass kicking LINK

T_M

So we remounted up the 3500+, and set to work on clamping down hard (we think we were a bit soft on first go), loaded it up to the brim with DICE and a nice dash of acetone.
You can see in the photo that we found ripping out some of the supplied insulation made accessing the nuts so much easier (note to manufacturer).

eva2000

ram and 74GB Raptor i guess

dinos22

eva i read about that iRAM stuff today once again.....that stuff would really kick the pants off drive dependant benches for sure

T_M

Now here come the sad part

Booted her up and all was looking fine (booted on my 2x256 on a fresh CMOS), temps were what we expected (still slightly higher than Venice though).
Then i switched it off and swapped the RAM for the 4000vx, meaning squeezing the RAM in and out from beside the insulation.



Powered it up, and all was fine so we incrementally increased HTT in steps of about 20 until something like 280 x 11.
Booted fine but BSOD on windows.
dropped it to 270 x 11 - still no windows
dropped it to 260 and finally got into windows.

rebooted and increased volts from our previous max used in the first runs to something like 1.86V (in mbm which shows less than actual).

Still not good so decided to try just some slow-ish PI runs using CLockgen to increase once in windows.
Think it was about 3000MHz.

Things werent going as planned and stability was worsening, so flicked on mbm to find temp was up to -10C!
Quickly we though "bad mount" so i very softly pushed on the tube to the side and immediately we saw it jump down to ~ -22C.
SO i wiggled it a bit more with the same result, however each time i let it go it jumped up higher and higher till we hit 0C.

Scragley feverishly set about tightening the nuts as best he could with the system off.

Hit the power button and for a breif second we got power, then nothing.............

I'll let this picture speak for itself.....

T_M

actually all this was done on a 120Gb PATA with a 40Gb PATA for a swap file

dinos22

what do you think did it in the end...........bad mount or too much voltage