Originally Posted by byteninja2
I want to cool my CPU.I want to overclock my CPU.I want to BEAT RECORDS!So, the current record is 8.8 GHZ on a FX 8100?
8.7ghz actually http://hwbot.org/submission/2294034_andreyang_cpu_frequency_fx_8150_8709_mhz
You probably won't beat it for a few different reasons, as I will point out below.
Lets beat it.So, first things first, what CPU?I have a little while, lets ay mid next year at the latest.Should I go with the FX8100?i think not.I betcha the record holder's 8100 was barley stable,
You will need an 8150 or 8120, but most likely 8150. The chip will need to be an early manufacture date (which are getting hard to find now days), as it is apparent that all the later ones do not clock as high. No other BD chips have the potential for 8+ it seems.
I probobly couldnt push it to 9 GHz.
I doubt you could get it to close to 8.2ghz, and I'm saying that because of the very few amount of chips that have.
I am thinking piledriver, they will probobly be able to hit that vital extra 200 MHz.
Pure speculation. The current BD can't clock very good, so I wouldn't get your hopes up for a architecture refresh, as they will probably be using the same manufacturing techniques as the current BD chips.
2nd, board.I am not a fan of Asus, actually, the oppisite.But, those Asus ROG boards look like some extreme overclocking boards.
If you're getting into extreme overclocking, and going for records, you can't let your hardware manufacture bias get in the way. Some boards are just better than others.
3rd, ram.I think I will use the same timings and Ram as the current record holder, A Data 4 GB.
He had 2gb of corsair. You can't really copy settings like that since lots of things can be different, like IMC capabilities, board capabilities, ram IC capabilities (even between identical chips).
Last, the most important part, cooling.This is the part were I am on a $600 budget.I think once you are into liquid nitrogen, its more about CPU Stability than about the cooling.
When running high volts on BD chips, stability comes from it being cold. Also, cold stops the high volts from kill the chip. Stability and cold go together.
Like, you might worry between H100 and NHD14, but sience almost all LN2 things give great cooling, who cares, really.
I'll let you find that out for yourself
. It would be boring for me to just tell you, given your preconceived notions.
What is a Rotary phase change/Phase change cooled system?How will it compare to a LN2 system?
It is a system that uses refrigerant liquid/gases to cool the chips via the Carnot thermodynamic cycle. Similar to HVAC equipment, but much colder. Cascades systems typically go to about -120C or so, but there are some that do -140C. They cannot keep up with LN2, they just aren't cold enough.Edited by just_nuke_em - 8/14/12 at 11:36am