boze wrote:
2) I guess I have to attribute this to the BIOS updates because nothing else changed, but idle CPU temps were drastically reduced with each of the two updates.
0307 -> 50.5°
0405 -> 39°
6066(beta) -> 26°
Or maybe this had something to do with the thermal paste on the heatsink warming up and creating a better contact surface? I really have no idea, but it seems cool as a cucumber now. I have the little utility open and it's ready 24° as I type this. ... is 39° normal for mobo temp??
Ah. I'm happy to know that you got these results. I spent a bunch of this week going round and round wondering why the new box I was building idled so much hotter than the last two I built, all with E6x00 on P5B-VM. Eventually caught on that the first two were bios 0207, and the third had 0307. I've gotten temps very similar to what you report, from 0307 & 0405. I haven't tried the 606 beta, but 0207 gave temps similar to or perhaps a little lower than 0405.
boze wrote:
I experienced several "Intel CPU uCode Loading error" messages during startup. I updated the bios from 0307 to 0405 and it still wasn't fixed. Updating to the newest beta version of the bios eliminated that error.
Odd. I didn't get this problem. I did of course get the "CMOS Checksum Bad - Overclocking failed" message at first POST after reflashing the BIOS.
And about those (huge!) temperature differences from one BIOS version to another... I can't decide whether the temperatures are really changed, or whether it's just that the measurement has been recalibrated. And I can't decide which truth I would find more disturbing. It's not just the temperature numbers reported inside bios setup that change. Numbers from ASUS PCProbeII, and Speedfan 4.31 or 4.32beta also change, by about the same delta-T, from one BIOS version to the next. And my understanding was that Speedfan read directly from the WinBond W83627DHG chip, so it's not clear how the BIOS would be intercepting that reading, unless maybe the bios preloads a lookup table in the WinBond, at POST? The sys/mobo & Core0/1 also changed, by about the same delta-T. HDD temps seemed to change also, but maybe not so much. But I didn't discern (by touching my scientifically-calibrated fingertip to the HSF base) any real CPU temp difference, from one BIOS to the next. And since ASUS make it difficult-to-impossible to revert to an older BIOS, it's hard to do careful 1-vs-2-vs-1 comparisons.
Anyway, the moral of the story is that anyone building on a P5B-VM should check which BIOS version they're on, before worrying about temperatures.
And regarding the conflict of cables vs. HR01...
boze wrote:
I'm still in the market for a passive HSF to replace the stock one but I think 120mm wide is going to be pushing it.
Take a look at the Akasa Evo-120. It might be just the trick for you. In the NSK1300 (w/ powerbrick & 200W DC/DC) I just built, the Evo-120 fit very nicely -- its fins sit about 8-9mm from my exhaust fan, and within 10mm of being perfectly aligned side-to-side. With this spacing/alignment, I didn't need any ducting, and there's just enough airflow escaping around the Evo120's fins to provide some breeze for the mobo. Your mileage may vary. MikeC points to a review at FrostyTech.com. The only source I could find for Evo120 here in USA was SidewinderComputers.com. I ditched their 'amber series' fan, and the aluminum shroud that held it on. My Nexus case exhaust fan is the only fan in this box.
Cheers!