Originally Posted by TonicX
Stress-test_002 no BCLK overclock has begun
1. Test trials continue on my render rig using the Ivy Bridge validated AIDA64 System Stability Test
2. Fans will be set 30-60 [40%power ate 30* scaling to reach full 100% fan power at 60*
3. MOW techniques will not be implimented (MOW = Minnesota Open Window)
4. The Case will remain stationary with both side panels off
5. the OC settings have been changed in BIOS to:
senerio B. Core speed 4900 Mz x49.0 , Bus speed 100 MHz Core Voltage 1.410 Offset Enabled
1. cinebench .................pass 9.78 (down from senerio A 9.90)
2. intel burn test................ success
Here's what I've learned from member posting concerning raising the BCLK
1. 100.1- 104.9 is a moderately raised BCLK and over 105 is more extreme dangerous [DEPENDING ON THE TEMP CONDITIONS AND COMPONENTS as with all OCs]
2. It is a viable option for benchmarking and short tests but should be avoided for Long tests because it is less stable.
The test is currently a the 1 hour mark for the BCLK @ 100 x 49 right now.
stay tuned for more updates! stay tuned! get it? Ha! I kill me. Official Corsair Hydro Series Club
5.5GHz Ivy 3770K
IF IT AINT BROKE THEN YOU DIDN'T OVERCLOCK IT ENOUGHT
Stress-Test_002 Has exceeded senerio A by one hour at 5:04:30 hours AIDA64 Stability Test: PASSED.
Im going to terminate the phase 1 of the EXPERIMENT
1. Since i was testing stability and the benchmark with Core speed 4897.8 Mz x47.0 , Bus Speed 104.21 MHz Core Voltage 1.410 Offset Enabled reported: 3/14/2013/ 6:35:37 AM - WARNING: HARDWARE FAILURE DETECTED! TEST STOPPED
2. and senerio B. Core speed 4900 Mz x49.0 , Bus speed 100 MHz Core Voltage 1.410 Offset Enabled exceeded this time by 1 hour and was manually terminated.
the hypothesis of: "Raising the BCLK instead of the multiplier is less stable for extended time-frame benchmarking" is supported.
I do not know what "HARDWARE FAILURE DETECTED"
i do know exactly what "TEST STOPPED"
1. I see why use of BCLK Frequency increases are undesirable and will limit the use of this adjustment.
2. If this adjustment is dialed into the mix i should be minimized and use for purpose of achieving frequencies between the hundereds numerical place of MHz i.e. 3900, 4400, 5100, also expressed as the Tenths place of GHz i.e. 3.5, 4.2, 5.5
3. Doing some quick math I see no reason to raise the BCLK on ivy bridge above 102.1
because at this point it achieves the equivilant of the next multiplier and then goes back to 100.0
3b. example: 49x multiplier with a BClk of 102 is 4998 and so it is pointless to go higher with the BCLK when It simple starts over at 50x 100.0
4. if i liked the speed, voltages, and temps at a clock between the pillar Multipliers say 4948 you obviously cannot achieve this with a bclk of 100.0 so in that case I would test it with some worst case senerio artaficial workload software like aida64, cinebench, Intel Burn Test, what have you to get a quick Idea if its good-to-go. but what really maters is the app I am using, Photoshop, After Effects, Cinema 4d, GRID, Battlefield3, Folding ect. is it stable. If It is I see no reason why not. OC that last ounce of performance.
5. I repeat my main discovery: NO NEED TO GO ABOVE 102 WITH THE IVY BRIDGE.
6. I will probably be Operating my Render Rig at Core speed 4900 Mz x49.0 , Bus speed 100 MHz Core Voltage. If I render with all 8 threads.
Thanks to all the OCN Crewmen!