Before I go much farther, I started reading other OC efforts. The guy at OCforums got his first one to 2.7 at 1.4 v cpu, but couldn't get it stable so he RMA'd it, but couldn't even get the second one to 2.5. I got that beat...lol. The Tomhardware says they got 2.6. Hmmmmm
The highest CPU temp I can find is 47C, but the caps are back into the 50's. NB at 37C and sb at 53C
Note: my AMD64 X2 5200 is now doing this same wu lot and coming in at 38 min and change. The Quad (at this level) is taking 41:06, 41:10, and the like.
I can guarantee it will try (after the well at Simap dries up..next day or so, it'll be dry, then run down the cache).. Oh heck, just wait and see....LOL If smoke isn't released, that is. Running my einstein total up got pushed back to I could run Malaria for Dogbytes in February, then I saw where the regular work at Simap was ending and they were going back to the "once a month" release of work. So, I figured I'd better "get some" while I could. Actually they have relatively stable run times so they're perfect for testing.
The stable run times I was talking about for the last 3 wus were 40:38, 40:35, and 4025, so they're pretty constant.
Changed FSB to 204 for 2.652 Ghz
FSB freq - 204
PCIE freq - 100
Processor freq mult - 13X
cpu-nb link speed - auto
cpu volt - 1.275
cpu-nb volt - auto
dram volt - 2.1
I just have to assume I'm about to hit a wall.... I just wonder what wall, and what'll happen.
[edit] It froze, rebooted and upped CPU V to 1.2875 Linux is now "recovering journal"
FSB freq - 204
PCIE freq - 100
Processor freq mult - 13X
cpu-nb link speed - auto
cpu volt - 1.2875
cpu-nb volt - auto
dram volt - 2.1
Looked at results and haven't found an invalid yet.
I'm finally there. Equal to the highest clock I have found by any OC review so far on air cooling (still looking). It's a 17.39% OC.
The last batch of wus came in between 39:37 and 40:01 .
I had to run out for some water softener salt, that's why I was gone so long.
I'm at 2.7 Ghz.
FSB freq - 200
PCIE freq - 100
Processor freq mult - 13.5X
cpu-nb link speed - auto
cpu volt - 1.3125
cpu-nb volt - auto
dram volt - 2.1
Mem freq 533/1066
2T, 5,5,5,15
SB at 51C
NB at 34.1
Ram at 34.1
checked each cap, ranges from 44-59C
max measured cpu temp 53C (using fluke 65 IR thermometer, which I can point directly at the cpu and heatsink pad.)
Now to watch the temps closely. But by "Gosh" I'm there. The others were unable to get a stable clock at 2.7, but they were running Vista. Not sure if that matters???
[edit] oops, got a freeze, bump CPU Volts to 1.3125
I'm finally there. Equal to the highest clock I have found by any OC review so far on air cooling (still looking).
[edit] oops, got a freeze, bump CPU Volts to 1.3125
So far what I see are people fixated on increasing the FSB. Why don't you try decreasing FSB to 193 and have a CPU Multi of 14 (193x14=2702)? If the FSB plays a part in the HTT clock, then at least for Athlon64 you can decrease HTT from the nominal 1000 (200x5) and there is still plenty of bandwidth... I admit that I have no idea about the NB linkage and if you'll then have to change that...or what effect FSB has on the memory... If it reduces the memclk like what it would do with K8, you could perhaps make up for some of the loss by going to slightly faster timings...
The lowest fsb option is 200. I've read many reviews where it craps out over 210. the Phenom has seperate NB, SB, and HT settings, and isn't dependent on cpu clock any longer.
to keep 2550 stable, I had to
)
to keep 2550 stable, I had to bump the cpu V back to 1.25. I see where tomshardware ran it at both 1.25 and 1.28, soooooo I hope it's OK
FSB freq - 204
PCIE freq - 100
Processor freq mult - 12.5X
cpu-nb link speed - auto
cpu volt - 1.25
cpu-nb volt - auto
dram volt - 2.1
Before I go much farther, I
)
Before I go much farther, I started reading other OC efforts. The guy at OCforums got his first one to 2.7 at 1.4 v cpu, but couldn't get it stable so he RMA'd it, but couldn't even get the second one to 2.5. I got that beat...lol. The Tomhardware says they got 2.6. Hmmmmm
The highest CPU temp I can find is 47C, but the caps are back into the 50's. NB at 37C and sb at 53C
Note: my AMD64 X2 5200 is now doing this same wu lot and coming in at 38 min and change. The Quad (at this level) is taking 41:06, 41:10, and the like.
OK, I did it. 2.6Ghz I put
)
OK, I did it. 2.6Ghz
I put the FSB back to 200, raised the multi to 13, and bumped up the CPU volts to 1.275
FSB freq - 200
PCIE freq - 100
Processor freq mult - 13X
cpu-nb link speed - auto
cpu volt - 1.275
cpu-nb volt - auto
dram volt - 2.1
Strangely enough CPU temp isn't rising as fast as I'd figured, still can't find a spot over 45C. still barely warm to the touch
RE: OK, I did it. 2.6Ghz I
)
But will it complete an E@H task ? ;-)
CU
Bikeman
RE: But will it complete
)
I can guarantee it will try (after the well at Simap dries up..next day or so, it'll be dry, then run down the cache).. Oh heck, just wait and see....LOL If smoke isn't released, that is. Running my einstein total up got pushed back to I could run Malaria for Dogbytes in February, then I saw where the regular work at Simap was ending and they were going back to the "once a month" release of work. So, I figured I'd better "get some" while I could. Actually they have relatively stable run times so they're perfect for testing.
Bikeman, look fast....your credit is 903,903
RE: Bikeman, look
)
Cheers
The stable run times I was
)
The stable run times I was talking about for the last 3 wus were 40:38, 40:35, and 4025, so they're pretty constant.
Changed FSB to 204 for 2.652 Ghz
FSB freq - 204
PCIE freq - 100
Processor freq mult - 13X
cpu-nb link speed - auto
cpu volt - 1.275
cpu-nb volt - auto
dram volt - 2.1
I just have to assume I'm about to hit a wall.... I just wonder what wall, and what'll happen.
[edit] It froze, rebooted and upped CPU V to 1.2875 Linux is now "recovering journal"
FSB freq - 204
PCIE freq - 100
Processor freq mult - 13X
cpu-nb link speed - auto
cpu volt - 1.2875
cpu-nb volt - auto
dram volt - 2.1
Looked at results and haven't found an invalid yet.
I'm finally there. Equal to
)
I'm finally there. Equal to the highest clock I have found by any OC review so far on air cooling (still looking). It's a 17.39% OC.
The last batch of wus came in between 39:37 and 40:01 .
I had to run out for some water softener salt, that's why I was gone so long.
I'm at 2.7 Ghz.
FSB freq - 200
PCIE freq - 100
Processor freq mult - 13.5X
cpu-nb link speed - auto
cpu volt - 1.3125
cpu-nb volt - auto
dram volt - 2.1
Mem freq 533/1066
2T, 5,5,5,15
SB at 51C
NB at 34.1
Ram at 34.1
checked each cap, ranges from 44-59C
max measured cpu temp 53C (using fluke 65 IR thermometer, which I can point directly at the cpu and heatsink pad.)
Now to watch the temps closely. But by "Gosh" I'm there. The others were unable to get a stable clock at 2.7, but they were running Vista. Not sure if that matters???
[edit] oops, got a freeze, bump CPU Volts to 1.3125
RE: I'm finally there.
)
So far what I see are people fixated on increasing the FSB. Why don't you try decreasing FSB to 193 and have a CPU Multi of 14 (193x14=2702)? If the FSB plays a part in the HTT clock, then at least for Athlon64 you can decrease HTT from the nominal 1000 (200x5) and there is still plenty of bandwidth... I admit that I have no idea about the NB linkage and if you'll then have to change that...or what effect FSB has on the memory... If it reduces the memclk like what it would do with K8, you could perhaps make up for some of the loss by going to slightly faster timings...
The lowest fsb option is 200.
)
The lowest fsb option is 200. I've read many reviews where it craps out over 210. the Phenom has seperate NB, SB, and HT settings, and isn't dependent on cpu clock any longer.
Hmm, Timing?? maybe that'll be next.
thanks
tony