Thank you kindly sir, I appreciate your time and others here as well.
At one point I did have DDR3 1333 4x2GB (mismatched pairs) running unlinked at 1667mhz @ 1.5v for a few weeks on the machine (just putting that out there). There were occasional issues causing GPU driver crashes, probably due to mismatched ram so I went to 2x4GB and everything is okay again no more nvidia driver crashes.
Studying my MCP temps over the years I always stressed about the "high temps" that are indicated by the system itself, normally "reported" around 72c idle, then up to 92c under load which is where I get nervous because that's hot as hell if it's celcius (but my system NEVER gets that hot as I'm feeling around the system for the source of this "near boiling" heat). So I was concerned when you speak of raising voltages further, HOWEVER...
I literally just read up on MCP (Southbridge) temps and someone said that there's a "glitch" to temp sensors in many cases when it comes specifically to MCP temp readings, where it's "locked" to providing readings in fahrenheit but will falsely say celcius. So in my motherboard's case that it reports 72c, it actually means 72 fahrenheit or 22.2 celsius!!!!! Now this makes more perfect sense to me and I can finally rest easy after 9-years of stressing about the high MCP temps hahaha. Not only that, but the MCP temp reading scare has kept me from pursuing overclocks beyond 25% all these years.
With that said, the x5470 idles at 47c and full loads at 82c which seems pretty hot so I don't really want to push for a further overclock on air.
Also seen your bios settings for GTLref being "default" vs. "auto". I do not have 'default' as an option so you see mine's all set to +00 instead of 'auto'
So when you say "bump your chipset voltage up few notches" which one of these options in the picture I've enclosed controls the 'chipset voltage'?Edited by PCgamer911 - 11/20/17 at 9:16pm