Overclocking Intel i5 3570k @4.8Ghz - Raystorm - direct die cooling (yes I am)

Ok just looking to get some feedback on this O.C.

4.8Ghz @ 1.472v [using offset] (max core temp (large fft's) 85c on hottest core(although it hovers around 79/80 mainly))

LLC @ 75%

everything else auto for now.

Normal use/games maxes it @ around 60c/70c (idle 30c @ room temp 25c aprox)

Some time ago I removed my IHS to redo the thermal paste (win) and my temps improved and I got to 4.5Ghz

But going any further need a large Vcore bump and so was a no go.

Being me, and after some reading I tried this:

1. Removed socket holding bracket
2. Removed IHS
3. Layered Tape on left and right of chip to just about same depth as die
4. used two small pieces of tape to hold chip in place on socket

Then, so that the same amount of pressure was applied by the waterblock I replaced the thin plastic washers on the mounting screws with thicker ones [white in picture] (about the same depth as the IHS) and then mounted the block in the usual way. screwing down in cross fashion and then using the adjusters to apply extra even pressure.


Booted, tested, Worked (just much cooler (around 60c max)) and then after some time I got to this:


Update 7-10-12 21:40 Started to manually input PLL etc and work down Vcore


(I do have lots of screenies etc, once I have sorted them out I will add more)

So whats your thoughts/ideas?

 

1.472v on Ivy Bridge, isn't it a bit much? I thought 22nm required less voltage.

 

On that last pic, Coretemp looks like it's showing an eleven degree spread across the four cores. Are you sure you've got a good mount?

 

Yeah it has always done that even before I started Ocing (BTW when it gets over 90c running up to TJMax they soon all line up). But I am testing various mounting pressures ATM. the thermal paste spread when removed does look even though.

I have got this down a little 1.448v still a WIP. which is why I was asking.

Intel still state that 1.52v is max Vcore.

[updated first post]

 

Without the heat spreader would it not be possible to get condensation forming on the bottom of the block?

Sent from my GT-I9300 using Tapatalk 2

 

Not that I have noticed from running it over the last week. (lots of testing/remounting etc.)

 

That would only happen with sub-ambient cooling. i.e. a water chiller.

 

Use "liquid metal" and replace the heatspreader.

This will ensure your chip remains protected and the temps will remain similar to as they are now.

 

I wonder what could be the temp if the HS wss made of sterling silver or silver plated?

 

This is what my IHS equipped 3770k hit @ 4.5 under smallfft and 2 670s under furmark in the same loop, that 300Mhz bump is not worth it with the risks of D2D and 1.4++

 

Core Speed Goes From 4499 MHZ at 1.2610V to 4800 MHZ at 1.448V after tweaking the cooling Sys.

 

Yeah I know what you mean but I should of said I have two 3570k's (Thank you work). This is my "testing" chip while I play around with cooling solutions/mods etc. So if I do kill it it's not the end of the world.

Just looking for the "Best Fit", I will decide what I run @ when I final build.

Any suggestions for a good "liquid metal"?

 

I think these new chips are bad overclockers, i can get a 60% overclock on my old i7 920!!

 

Bad how? Just different its all about the math not the gigglehertz. Get use to it as they shrink in size.

Sent on my CM10 JB powered i9100 by TapaTalk 2

 

Hello!
I interested in what kind of heat conductive material used between the CPU core and the water block list? Thank you!