RESULTS (Good news/Bad News)
A lotta messin a roun today, lots of oppertunity for this thing to gag. The GOOD NEWS is, still at ZERO on the post-fail since BIOS update, (despite at least one INTENTIONAL attempt to make it post fail) And other than that one brief trip up to 65C when I'd put the VCORE voltage at +0.250 the All temps even under some load testing have been in the 41C or LESS range. And no instability, spontainious re-boots or issues getting in to the BIOS. That's the good news. THE BAD NEWS is SOME the results of todays testing has been not exactly consistant wth common logic. (EG: boost a multiplier, speed goes down)
Here's wher I left off: VCORE voltage at +0.125, cpu clk ratio at 18.00 (=3.60Ghz), Net VCore 1.54-1.6+v and temp centering around 39C in BIOS IDLE, actual Clock speed 3.78.
QUICK FOOTNOTE UP TOP. I use "VCORE" voltage (net) and "CPU" voltage (net) interchangably.
Change 1: Cut VCORE v by half to 0.075, no effect on ClockSpeed, remain 3.78, Net vcv down to 1.488 and temp 37C.
Change 2: Cut VCORE v to +0.050v, CS still showing 3.78 effective. net VCV down to 1.34 - 1.460, temp 32C
Change 3: Cut VCORE v to +0.025v Still 3.78 effective, net VCV 1.34-1.44, temp eged up 34C ( I write it off to time running)
Change 4: Cut VCORE v to 0.000v STILL SHOWING effective CS of 3.78, this is the first big inconsistancy. The last thing I did yesterday to get it UP to 3.78 was the VCORE v treak from +0.025 to +0.250, now I've dropped it all the way back to LOWER than it was and it's still showing 3.78Ghz effective. Temp now 30C and net CPU power is at 1.29 - 1.41v.
So temp and NET CPU voltages are both responding a logical manner, the CPU clock speed is not.
Change 5: CPU clock ration up to 18.5 (=3.88 antisipated) actually half expecting post-choke here, nope, post good! Reslutant CS now 3.70 (VCORE v still at +0.000)
Change 6: VCORE v up to +0.050 No change 3.70 (multiplier still set to 3.88)
Change 7: NB freq up one tick to 2200Mhz, no apparent effect. Put it back to Auto, (2000Mhz)
Change 8: took CPU multiplier back down to 18.00 (3.60) noticed VCV at -0.050v not sure if I did that by mistake or some how it something else auto-tweaked it. Sorry. ANWAY kicked it back up to 0.000 -restart- aaaaand BACK to 3.78Ghz CS.
TEMP 29-30C net V 1.32 - 1.42 Freq X18.00 (supposed limit 3.60)
Boot into windows.
CPU-Z advs: Core Speed 1470 / Mult X7.0 (7-19.5)
Core Voltage 0.936, 0=32c, 1=31c and 2=10c
Leaving CPU-Z open, time for some load testing. (In the form of FLV to AVI conversions)
1st conversion ran about 4 or 5 minutes. Temps peaked out as follows: 0= n/c(32) 1= 39-40c and 2= 26-27c
Watched some vids for a bit, 5-6 minutes. Temps 0 = 32, 1= 35/34 and 2= sub 20c
MORE conversions, about 10-15 minutes worth. Temps peaked at: 0=32c 1= 41c and 2 = 30c
About midway through this test I dailed the Case Fan speed dial all the way back. (virtually silent) to see, well you know.
as you can see, no big effect on the temps. Granted these are pretty short load tests, But I'm keeping in mind when I had the VCORE v at +0.250 the CPU temp shot up to 65c almost instantly. GETTING THE FEELING there's a very fine line here between just fine at 41C and OVER the (60c) LINE.
So now I
Change 9: kick the VCORE V up to +0.025, (restart) in BIOS IDLE now, temps CPU=32c and sys=32c net CPU v 1.28-1.39v (see bold above)
CRanked the case fan dial back to max, over the course of a bout 3 minutes pulled the SYS temp down 1C CPU temp n/c.
Still at 3.78 effective, mem is 1959
Done for the day. Shutdown.
Hopefully this was of use(or at least some MEANING) to someone.
So wadda y'all think? ADDING a good after mkt CPU cooler should I be able to get, a stable 4.0Ghz out of my 6100, long-term, without damage or throttling?
Edited by NoNewEgg4Me - 5/19/13 at 9:34am