I think it is generally agreed that the quality of the average waterblock review is pretty low - due to a variety of factors, the results are not scientific (read: repeatable). I don't see how looking at individual results is better than seeing how all the blocks compared using the same testbed using laboratory standards.
That is not to say BillA's graph is the be all and end all, since different designs work better on different sized heat sources (some blocks cool a small die like an AMD XP better, whereas some excel on larger heat sources like peltiers or AMD64/P4 heatspreaders) and some designs cool "hot spots" better, allowing for higher stable overclocks. However, I believe it is fair to say that his results are a better gauge of performance than a review where the quasi-reliable insocket thermistor is the only judge.
__________________
If not, why not?
|