MikeTimbers
31st October 2002, 16:54
Firstly I would like to thank Tom for supplying the chip. I'd also like to thank my mother and father <sniff> and my next-door neighbour's cat <sniff> ... <wipes tear from eye>
Anyway, no piccies cos no digicam but the chip is AIUCB 0240 so week 40. Updated the BIOS of my K7S5A and checked it re-booted with old chip. Then shutdown and disconnected the power cable.
I removed my old XP1500+ which has been limping along at 1380 since the Vcore dropped from 1.75 to 1.72. Anyway the XP1500+ was running around 122000 structures per day at DF.
So I cleaned the traces of AS2 from my Alpha 8045 and then a tiny bead of AS2 onto the XP2400+ before spreading it out with a razor blade. Then I put the new chip in the socket, re-mounted the Alpha and without a fan (!) checked to see if it would post. Yes!
XP1800 (mobo defaulted to 100 FSB).
Switched off and power cable out again. Re-fitted my Mechatronics 80mm back onto the Alpha (sucking, of course) then re-fitted the duct to the rear exhaust.
Switched on and went into the BIOS. Changed the FSB to 133 and re-booted.
XP2400+
W2K loaded flawlessly, DF started, everything looked sweet. Loaded Speedfan and voltage steady at 1.63v, temps around the same as the XP1500 had been. Using Speedfan, I selected the next available FSB - 138 for 2070. No problem. Then 142 for 2140. No problem. Then 146.7 for 2199. Still no problem. CPU temp (this is all with DF running of course) showed 46C. This is with the two Mechatronics on my duct at 7v, remember.
To go to 150FSB, required changing the BIOS and a re-boot because at 150 the 1/5 divisor kicks in and changing while running will lock the mobo every time. So, POST at 150? Yes!
Then BSOD loading.
Well, I can now do one of three things. Change the voltage by closing the L11 bridges to get 1.75 or higher. Cut the L3 bridges to change the multiplier to 16x. Or wait and let it burn in for a while at 2199.
I'm being lazy and tryig the burn-in for a while. 10% at default voltage for these early steppings seems conservative but I really don't want to blow up this chip.
As for DF...
how about these numbers:
------------------------------------------------------------
Distributed Folding Windows dfGUI v1.9 Benchmark
Sample Size : 7950 structures over 3983 seconds.
Protein Size: 129AA
Structures Per Second: 2.00
Structures Per Minute: 120.0
Structures Per Hour : 7198
Structures Per Day : 172757
OS : Windows 2000 MHz: 2199
CPU: AMD Athlon(tm) XP 2400+
------------------------------------------------------------
Not too shabby, eh?
Anyway, no piccies cos no digicam but the chip is AIUCB 0240 so week 40. Updated the BIOS of my K7S5A and checked it re-booted with old chip. Then shutdown and disconnected the power cable.
I removed my old XP1500+ which has been limping along at 1380 since the Vcore dropped from 1.75 to 1.72. Anyway the XP1500+ was running around 122000 structures per day at DF.
So I cleaned the traces of AS2 from my Alpha 8045 and then a tiny bead of AS2 onto the XP2400+ before spreading it out with a razor blade. Then I put the new chip in the socket, re-mounted the Alpha and without a fan (!) checked to see if it would post. Yes!
XP1800 (mobo defaulted to 100 FSB).
Switched off and power cable out again. Re-fitted my Mechatronics 80mm back onto the Alpha (sucking, of course) then re-fitted the duct to the rear exhaust.
Switched on and went into the BIOS. Changed the FSB to 133 and re-booted.
XP2400+
W2K loaded flawlessly, DF started, everything looked sweet. Loaded Speedfan and voltage steady at 1.63v, temps around the same as the XP1500 had been. Using Speedfan, I selected the next available FSB - 138 for 2070. No problem. Then 142 for 2140. No problem. Then 146.7 for 2199. Still no problem. CPU temp (this is all with DF running of course) showed 46C. This is with the two Mechatronics on my duct at 7v, remember.
To go to 150FSB, required changing the BIOS and a re-boot because at 150 the 1/5 divisor kicks in and changing while running will lock the mobo every time. So, POST at 150? Yes!
Then BSOD loading.
Well, I can now do one of three things. Change the voltage by closing the L11 bridges to get 1.75 or higher. Cut the L3 bridges to change the multiplier to 16x. Or wait and let it burn in for a while at 2199.
I'm being lazy and tryig the burn-in for a while. 10% at default voltage for these early steppings seems conservative but I really don't want to blow up this chip.
As for DF...
how about these numbers:
------------------------------------------------------------
Distributed Folding Windows dfGUI v1.9 Benchmark
Sample Size : 7950 structures over 3983 seconds.
Protein Size: 129AA
Structures Per Second: 2.00
Structures Per Minute: 120.0
Structures Per Hour : 7198
Structures Per Day : 172757
OS : Windows 2000 MHz: 2199
CPU: AMD Athlon(tm) XP 2400+
------------------------------------------------------------
Not too shabby, eh?