September 2000

Introduction
The first submersion project, Mission: Submersible, was something out of the ordinary.  It was the first time 2 extreme elements, Fluorinert and Liquid Nitrogen, were used in such experiment.  I know Cray used fluorinert before but he didn't cool this liquid to subzero levels much more use LN2 to supercool it.  His experiment was done in a very controlled environment with all the cool stuff and gadgetry.  Mine was just done in the backyard.  It was an ambitious experiment and as everybody said, a very expensive one.  Especially for a simple overclocker like me.  Overclocker....simple?  Mad is a better term.

After the experiment was published in the site, we were blown away by the publicity it generated.  There were a lot of praises, suggestions, and a few negative remarks.  The PC community round the world were very amused that such an attempt was done.  All walks of life came to see and read the article.  Emails flooded my mailbox by the hundreds just to say something about the experiment. 

A lot of question were asked as well.  The most common ones were: 

• Why didn't I check the pour/freezing point?  For the very simple reason that when I bought the fluorinert, I wasn't planning of supercooling it with LN2.  My goal initiallly was just to submerge the rig and probably place a cooling element from an old refrigerator or use freon and a compressor to cool it.  But because hundreds has already done that sort of supercooling, I started thinking of a different method.  Thus, came the idea of LN2 as the supercoolant.  Also as I've said before I always wanted to play with that stuff and what better timing.
  

• Next was why FC-40?  Not the other type of fluorinert that has a lower pour point.  Again, same reason as the latter + the fact that when I asked for the liquid, the only available type at the time was FC-40.  Any other type would take 6 weeks to arrive here.  I cannot wait anymore.
  

• The next question some people ask is why a Celeron?  I chose the Celeron as my guinei pig not for the simple reason that it was cheap in case it gets stuffed, but mainly because of its 66 mhz front side bus.  It gives more room to overclock as opposed to a Pentium III which already has 100mhz as its front side bus. 
  

• Some other comment was that I could have easily bought a 1 Ghz CPU for the money I've spent on that experiment.  My answer is YES. I could have done that, but where would be the fun and challenge in it?   I have bought P3 600's, 650's and 700's in the past and have overclocked 2 of them rock solid to more than a Gigahertz, here and here, so I'm not after that.  What I am interested and curious at is what supercooling could do to the whole setup while fully submerged in fluorinert.  Will it boot?  Run programs and extensive benchmarks? Most importantly will it overclock much much better? 

In the last project we were able to take the stubborn Celeron 366 to 650mhz , a chip that only overclocks to 550mhz at 2.3v with a 52w peltier before the submersion.  That's 300mhz overclock or equivalent to 83% above spec.  The experiment experienced some disaster but overall it was a success.  We encountered some problems which was expected as it was the first time such a thing was done.  So no guide to follow or theories to rely on as there is no real data to review since it was as Ived said never been tried before.  But now we've located the problems and after studying the experiment over and over again we're quite confident now in doing the sequel to it.  Because we didn't fully maximize the full potential of a supercooled fluorinert....  With the help of the thousands of suggestions you've shared to me, I NOW GIVE YOU.....