My friend and I recently invented a strange new cooling system, and wanted to give it something to show off with. We chose a Prescott 2.8E, and an Abit IC7 motherboard. We're using PC4200 OCZ memory.
Unfortunately, there doesn't seem to be very much data at all on the Prescott. We know it idles ~50C in air, and ~70 when overclocked + cooled with EXOS (courtesy hardOCP).
We managed to get the 2.8 to 4.13GHz, with a CPU temp of 47.5C. That's 147.5% faster than stock.
Is this pretty good? We're using the 14x multiplier, so theoretically with a 3.2 we'd be at ~4.7-4.8GHz, and no, we're not using liquid nitrogen or dry ice, OR a vapochill system.
My question is, are these results promising, or run-of-the-mill? I'm trying to decide if we should continue to refine this setup before we go public with it, or if it's nothing special and we should can the project where it stands.
Results sound pretty good to me, but that depends on the cost of the cooling solution. I can't really answer your question except to suggest you try more than one chip before "canning" or diving into producton. Every chip is different. If you have a great OC'in chip, your results might be soso. if you have a poor OC'in chip, you might have discovered gold. See what the best normal air cooled OC is on that chip and motherboard. Is it way lower? Now weigh in the cost of your cooling solution. That's my opinion anyway.
yeah, there are way to many variables to be able to give you an answer. youd need to test over several systems with varying platforms, such as motherbaords, multiple prescott chips, maybe northwood and even amd chips to see how it fares. and depending on your apparatis, you have to figure out what you maximum dissapation is and push it to that limit.