Project : Little Stealth II

As you can see, it's a Socket A block. After shearing the lugs off my last motherboard, and then having a really difficult time selling that  motherboard because of that, I decided I would never even attempt to mount a  waterblock off those lugs ever again. Chip came out with his Socket A blocks shortly after I placed my order. Luckily the silver blocks were on backorder, so  I just asked him to change my order. Socket A blocks use the 4 holes that are placed around the CPU socket. These holes are also on Pentium 4 motherboards,  it's actually part of the ATX 2.03 standard. This isn't the only waterblock  using these 4 holes, all 3rd generation waterblocks are using them. And most second generation blocks can be adapted to use them. But this is where we come to the complaint I have with this block. All other Socket A waterblocks, and even 1/2 of the Socket A adapter plates for older waterblocks come with SPRINGS. The springs go between the nut and the top  of the waterblock. So when you tighten down on the nuts, you can not only see how much force your putting down on the core, the springs are specifically selected to exert the correct amount of downward force on the core. This has 2  benefits, first of all, the waterblock is self-leveling, you don't have to worry  about it being cocked off to one side and not sitting flat on the core.. Let me tell you, it's HARD to tighten 4 screws evenly all at the same time. And the  other benefit is that you won't tighten too far and crush/shatter the core of the CPU. Now I discussed this with Chip for ages, I tried to make him see it my way. No dice, he's dead set that his no spring method is better than the spring method. All in the eye of the beholder I guess, as you can see, I did mount the  block properly without springs, but I also ordered a set of springs from HighVoltagePC, and I  intend to install them when they arrive.

So how does the waterblock perform? Awesome.. I now have a core  temp probe, it's an Asus P2T, this is a thin thermistor that you can place  anywhere, even between the ceramic part of the CPU and the waterblock (not  between the silicon part and the waterblock tho! heh). I trimmed off the excess  plastic around the tip of this probe, and placed it directly against the base of  the core (the actual silicon die of the CPU). This is how I monitored temp on my old Athlon classic when I was running a K7M motherboard, it's much more accurate than the socket temp probe, and updates alot faster too, you don't have to wait  for the bottom of the CPU to change temps. So now I can see both temps in MotherBoard Monitor, and I can see that the socket probe is indeed 10f too high  like it had been rumored to be, tho it tends to vary, it can get even farther out at different temperatures. Ok.. So, just to establish how much heat I'm feeding this waterblock, I have a green core TBird 750 (green means it has  aluminum interconnects, not copper like the blue ones have) at 966MHz, and 1.9v,  that's a modest 62 watts of heat. Here are my temps:

That about wraps this article up,  If you have Questions please post them on the ProCooling Forums (recommended), or Email UnaClocker directly.