Originally Posted by Benjiw
You actually did watercool the Kitty? I am struggling with the idea for some time now, you think it is worth it?
How are the vrm temps now? Most notable plus is going to be clean looking'ness. I just hate the Noctua ruining the look of my build.
Originally Posted by dmcl325i
Im not as think as you confuse i am
I just wasnt sure where people were talking about tmpin1 being at. Now i know the location i need to cool i will cool it lol
You can see though all my other temps are pretty good so that little weeny h55 is doing the trick with the sp120's on it. I just need to figure a way to cool the actual socket/mobo around or behind it. I think a fan of some sort on the back will do the job. The 140mm fan i stuck on there sucks for that, it doesnt flow well being so close to the mobo, it must be designed for unrestricted flow. I might re-run same test with one of the boggo corsair 120mm case fans in its place tomorrow and see if it improves.
I may be back up to 4.8ghz yet.. Although i think i have more or less proved my point on the single slot weeny little h55 cooler in an sp120 sandwich. Theres even a little air leakage between the front pusher and the rad, might make up a little gasket or something so its forcing all the air through the rad with no leaking around the edges.
Hmm. You should almost go promoting hardware, I think many people will buy it.
Stop putting mediocre stress on the chip and say its stable.
Low end AIO with "okay" fans isn't gonna cut it.
About the gasket, they are being sold. I've had them. Here's a linkhttp://www.frozencpu.com/products/18046/ex-rad-469/Phobya_120mm_Radiator_Gasket_-_10mm_38335.html?tl=g30c637s162
Originally Posted by mus1mus
I don't know what's harder to get a hold of.
1. IBT AVX at the first page.
2. Very High - not HIGH
3. Your claimed temps are the minimums.
4. Stop talking like you are stable. You are not. You wanna prove it? RUN Prime Blend for even 2 Hours and/or IBT AVX as mentioned, 50 Passes.
People have been so kind to point you in the right direction and you keep ignoring that fact and injecting your point. It's really pointless for people that have been doing this longer than you have your system.
Lastly, SP120s suck.
I'm really amazed how many people got involved in this since the last few days. I somewhat started with my KIND advice. But now I'm really done with it.
If OP decides to actually do the proper tests and encounters problems (which he WILL), I won't laugh and will also be happy to help. No problem.
That's what makes OCN special IMO. the knowledge.
You are very correct BTW.
Originally Posted by dmcl325i So you say 50 passes, others say 10 or 20
.. Im not going to deliberately run the system in to the ground because one person wants to see this stress test and someone else wants to see that, etc.
Apart from needing to sort out something more permanent for the rear of the mobo. All other temps were fine as said i dont know how many times previously.... You can clearly see min and max temps. With the exception of tmpin1 which is socket temp, everything look fine. And that is socket temp and not cpu temp... Cpu temp sits around or below 40c.. And thats not minimum, minimum at idle is in the mid 20's usually for cpu.
Currently eyeing up server fans and projector fans. Not looking for gimmicky junk for cooling socket and possibly nb/vrm sinks if i change from the current little fan covering the two heatsinks. Im looking at fans ideally pulling in excess of 1-2a that i will need to power via ide or sata plugs from the psu. Dont intend to fart about with cooling for socket/nb/vrm areas.
. The temps are low because you don't stress hard. The idle is to be ignored anyway, because it is way off. It's not even close to being accurate. But that's no problem, load temps are what matters anyway. Socket and package are BOTH cpu related temps. It can be confusing I know. Cpu cooler also has an effect on the socket temp.
I told you a few days ago to do prime95 smallFFTs for like 10 minutes and/or IBT (avx) very high for 10-20 runs. That is during the process of upping your overclock. To test for actual long term stability one can even run IBT for like 500 runs. If the system is stable it won't kill it. But not testing at all will ruin thing in the long run.
Originally Posted by Benjiw
Oh boy! This is fun to read but I bet it will be gone from existence soon.
I forgot to do some overclocking this weekend tbh I haven't used my PC since last week. Things in my flat still don't have homes and I collapsed the other day due to doing too much. Meh need to get my student finance sorted and get that i5 rig and do some apples to apples.
As to our new friend, Get IBT AVX and do some proper testing, I know you think a lot of what we are saying is nonsense but it really isn't and no amount of imagination and misreading temps etc will escape the stone cold facts that you aren't stable, you aren't testing correctly and you're pumping too much heat into that tiny rad which once you start actually doing some real stress testing, you will understand.
I'm finishing with this post Benjiw.
I can quote the whole conversation going on between the new guy and the old gang but it is pointless as it is.
It's all been said.
Then 120 rad is NOT going to cut it. Well. 4.5-4.6 max, yeah sure. More? Nope
Even my old h100 which performs at least 1.5 times as good as his h55 struggled to keep my chip cool at 4.8. Enough said.