-
Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!
10 X better TIM
- Thread starter MILLTHERM
- Start date
- Joined
- Apr 14, 2003
- Location
- Brooklyn
if it really does get to the market and is cheap and becomes widly used by q1 2004 then that would be awesome
- Joined
- Mar 3, 2003
- Location
- New Jersey
looks awesome! more for the ppl who don't change their setups often obviously
- Joined
- Nov 17, 2002
- Location
- New York
I see limitations, ie...you bond/weld your heatsink/waterblock onto the chip, chip dies, how do you get it off. Do you throw away your heatsink/waterblock with the dead chip?
- Joined
- Feb 7, 2003
no, youd bust the chip off and sand smooth at the worst case scenario.
- Joined
- Mar 28, 2001
- Location
- The Netherlands
Mmm....ya'll leave the pins in the socket, right?
Or do you guy's buy a new socket and solder it yourself? (I probably could...but do you?).
I mean how do you get the lever up w/o removing the HSF?
Or do you guy's buy a new socket and solder it yourself? (I probably could...but do you?).
I mean how do you get the lever up w/o removing the HSF?
- Joined
- Nov 17, 2002
- Location
- New York
Probably not worth the trouble... not right now anyway.
- Joined
- Mar 28, 2001
- Location
- The Netherlands
MILLTHERM said:
Mmm I once bought an Asus P5B-A (not sure about the type..it was a SS7 ATX MOBO) after I had some good experience with AMD CPU's/MOBO's (Before, my first PI (K5 PR75), I was P'd off at AMD) I decided to sell my AT MOBO (P5A orso) and K6-II(?)400, well after it ran for 5 hours (Q3) it hung...tried all I could think off...took it back..guys at the store wanted to have a look....bad PSU...new PSU...still the same...took it back again....NO WAY they could get the HSF off...uhh w/o using a flathead screwdriver.
While I bought a PIII 733 back then and let them have the MOBO+CPU, I'm sure they've cut some traces on the mobo, and if a normal thermal pad can hold a HS like that on the CPU, I'm sure it takes MUCH more force to get a CPU that's welded...
Never mind...I think you're thinking of AMD CPU's...then you're right, and you might be able to pull it off with the IHS Intel CPU's...although I heard it's hard to get the IHS off.
- Joined
- Mar 8, 2003
How would this compare to the carbon black tim that the engineer in the east made (near new york-- i dont really want to look it up- recent though)
- Thread Starter
- #14
I assume that you can solder direct to the silicon which would be a major cooling advantage.
The obvious strategy would to put on the largest heat sink possible, Thermalright or whatever. if the fan fails change the fan that would about the only option.
Clamp your heat sink down and light it off and you are set' for good
This stuff burns 15-20 ft per second as I smoke I am not sure I would want a couple of cases around I would probably end up welding myself to the wall
The obvious strategy would to put on the largest heat sink possible, Thermalright or whatever. if the fan fails change the fan that would about the only option.
Clamp your heat sink down and light it off and you are set' for good
This stuff burns 15-20 ft per second as I smoke I am not sure I would want a couple of cases around I would probably end up welding myself to the wall
- Joined
- Dec 29, 2000
A good application would be to bond the heat spreader on the P4 to the silicon. AMD could use this to. If we had a decent heat spreader even, permanently attached, that would go a long way towards making our heat sinks more effective. I would try bonding a 1/4" thick copper plate and then adding my water cooling on top with good grease.
Altough, as long as AMD chips stay cheap (I paid $50 for my 1700+) I suppose I could just build a water block and bond it on. But if taking it off kills the CPU it would make you decision to bond pretty crucial, $50 is still $50. I wonder if the CPU maker would sell them pre-bonded to their stock HSF's that would be bad news.
O
Altough, as long as AMD chips stay cheap (I paid $50 for my 1700+) I suppose I could just build a water block and bond it on. But if taking it off kills the CPU it would make you decision to bond pretty crucial, $50 is still $50. I wonder if the CPU maker would sell them pre-bonded to their stock HSF's that would be bad news.
O
One of the things most folks don’t realize is modern CPUs have multiple heat sources. Depending on the task, the CPU will heat up in different areas. This causes the core to flex. Imagine the top of the core as a little like the surface of a waterbed. Here’s a paper from Intel that illustrates the concept.
If a heatspreader is soldered directly to the core, the core will eventually crack from the stress.
If a heatspreader is soldered directly to the core, the core will eventually crack from the stress.
- Thread Starter
- #18
Maybe thermal stresses are lower due to lower temps.
the TIM could expand and cotract with the die.
another concern about direct die come from those that tried lapping the die and killed thier chips due to damage to the diffused silicon junction, probably only a few thouanths thick.
Similar threads
