AFAIK no way within HWiNFO to save graphs, you will have do as f1ll highlighted in his post
Originally on 0902 / 1002 I had used SOC: 0.950V for lengthy back to back stability testing with 3.8GHz / 3200MHz C14 1T. I found at low loads / idle occasionally rig would just go code 8, upping to 0.962V solved issue. As posted before I believe this happened on low loads / idle as voltages may just be dipping lower compared with when rig loaded.
When I went 0079 I found 0.950V did the same but quicker, 0.962V was perfect. As posted by me earlier my take on this is 0079 and newer have better RAM bench result in AIDA64. So I reckon this extra "efficiency" showed quicker 0.950V was too low compared with UEFI 0902 / 1002.Warning: Spoiler! (Click to show)
I do believe my voltages are spot on for what I set rig up as, but will bare in mind that I may need a notch or 2 higher on potentially VCORE/SOC.
Since last night going to 0081 which basically is how 0079 was setup (ie I was using ProcODT as 53.3Ω), rig has been behaving
. I'm today only keeping to using my 3.8GHz 3200MHz C14 profile and not doing any changes like yesterday or unplugging power. So far this mornings testing of normal usage / rebooting / shutdown and up, all has been well
, keeping detailed log what occurs today.@elmor
Thank you for viewing and commenting
. As highlighted to finalheaven all is well today
. Not going to "tinker" today, just to see if it fall over in normal usage like the last occurrence posted on 0079. I will see if AMD CBS settings are reset upon "issue", if nothing happens today then tomorrow I will "tinker" to make rig fall over and get AMD CBS screenie.@Tyrluk
Nice to read another CPU got you 3200MHz, shame not as good for OC, mine were very similar on that aspect so easy choice which to keep. Even X CPU does not guarantee better IMC from what I have read from shares by owners. I think I linked a post from XS forum before where Fugger
who has used extreme cooling on CPU stated his X CPU did not do same RAM MHz as 1700
Perhaps it's not IMC, as RAM speed affects "Data Fabric" clock perhaps it is something to do with that. Some CPU handle better others not. Again no idea just a thought bouncing in my head. Perhaps as FW improve it will get better or perhaps it won't. Only then we'll have better idea.
Thank you for sharing you tried another RAM kit. Like I posted before I don't think it's RAM / mobo, more like FW and/or CPU.@Masterchief79
3.6GHz OC I didn't try. As basically when reading The Stilt's Anandtech thread
he said this:-
One of the major down sides of the "OC-Mode" is that upon activation both Turbo and XFR will be disabled as well. Basically, this means that unless you are able to reach at least the default MSCF / XFR frequency on all cores, then you will essentially be losing single threaded performance compared to stock configuration.
So I went for min 3.7GHz ACB as that was only 50MHz below XFR and 500MHz above stock ACB.
1st needed ~+50mV offset to reach 3.7GHz stable in 48 loops x264. HWiNFO showed MAX CPU Core Voltage (SVI2 TFN) ~1.244V, I may have a ProbeIt VCORE reading on my notepad. 3.8GHz was ~+137mV, this passed a lot of back to back testing of Y-Cruncher / x264 / f@h / RB. 3.9GHz was ~+181mV tested with LLC LVL3 at the time = ~1.465V on ProbeIt VCORE, passed 10 loops x264, due to VCORE requirement and gaining more knowledge from others I backed off from testing more or going higher. Just as added info, 3.9GHz +181mV LLC: [Auto] was ~1.405V on ProbeIt, which would fail x264 loop 2 or 3 with WHEA errors.
2nd I have not tested for 3.7GHz. 3.8GHz is +162mV, ~1.380V on ProbeIt VCORE, HWiNFO shows MAX CPU Core Voltage (SVI2 TFN) ~1.356V. 3.9GHz is +250mV, ~1.475V, tested only on x264 20 loops and then discarded profile as VCORE requirement too high for level of stability I want.
All results above are LLC: [Auto] except the one instance highlighted.
4.0GHz it will bench in CB with +250mV offset, LLC: [Auto]. I can not go any higher on clock with that offset for CB bench. Below are results with CB PB in UEFI enabled.UEFI 0079 (ProcODT as 53.3, what is default in 0081) (Click to show)
UEFI 0081 (Click to show)
Edited by gupsterg - 4/13/17 at 4:58am
The 1786 result in below screenie is same setup but with CB PB in UEFI disabled.