Singlecore hwbot rankings damage with AMD

[TiN]

It's never late to restart AMD benching a bit Last AMD system i've ever benched or used was X2 3800+ Socket 939 rig. Then Conroe released, and I forgot about AMD stuff for a looong time

But looked few days ago a $35 price on Sempron 140 and rankings at hwbot at 1xCPU wprime I've decided to go some cash for it

Bought 2 CPU's, just for me and my teammate T0lsty from Xtremelabs.org , then we borrowed LN2 pot, 50L of LN2, Gigabyte 790X-UD3P mobo with some drams. PSU is my standard Enermax Revo 1250W, temp meter - Fluke 87V DMM.

In final we got this:



Mine 0664 CPU is ~150MHz weaker. That was yesterday.

That's sure not enough for me , so i took off IHS from sempron, just like case with :\:\ cellie . BTW AMD still don't use solder to fix IHS on die, so it's dumb easy, just accurately cut resin and ihs cap is off...





In rig:



Tests done with next specs:

Sempron 140 IHSless
Xtremelabs.org LN2 MAGNUM LE CPU pot
2x1GB Crucial Ballistix drams
ES 8800GTX vmodded/flashed to Ultra
2 x Gigabyte IRAM with 2x1GB DDRs for PCmark
Kingston X25-E 32GB as bootdisk
Enermax Revo 1250W



CPU w/o cap pushed very little compare to normal package. About 50-70MHz more. I was expecting better results

Results with IHS-less CPU:

CPU-Z valid 5710MHz



With IHS was 5655MHz.

PCmark2005 12256

5th total 1xCPU PCMARK, 2nd Sempron 140 rank


Wprime32 - 25.737sec
4th total 1xCPU PCMARK, 4th Sempron 140 rank

After 3hour session I've disassembled system and checked CPU die.
It's ok after 2.5kg copper pot torture on it!



So if you think about AMD IHS removing, forget about it. It's not worth that risks of damaging CPU die by evil cooling devices.