Sweet mother of god, New swifty heatsink

[BillA]

many Thanks roscal, will attend to it

CrashOveride
no, unless your ballpark is rather large
-> however the difference between devices should be similar

but here we may be ignoring things such as procedures, number of trials, etc.
these are not trivial and may influence the outcome
a very large chip mfgr requires 10 trials and a statistical analysis of the results

note also that some testers cite the best result,
while others (such as Swiftech) cite the mean from multiple tests
- the mean will always be lower of course

be cool