Average AR for a WU ?

[sean2002]

What is the average AR for a Seti WU ?

 

[Smoke]

0.417

 

[sean2002]

Greg, please you got that number from the last three WU I uploaded. If you didn't, this is real strange because the last 3 WU I completed were 0.417

 

[Smoke]

If you are interested in testing your system to see how it stacks up against other rigs you can use the "test WU" available from Ars Technica.

Check out their site that tells you all about this: Ars Technica Test WU

 

[Orange Kid]

0.417 is the 'usual' AR for most Seti wu's...............i don't think anyone has figured out a 'real' average AR from a users standpoint

 

[Poof]

Back well over a year ago, there was a thread at Ars discussing the merits of not using a VHAR WU for a benchmark (it was a 6.7xx at the time for the 2.x & 3.0 clients, IIRC). The discussion revolved around the fact that the best benchmark WU would be one where all or nearly all of the SETI scientific analyses would be taking place (VLARs & VHARs don't have this), in addition to the most prolific WU.

A number of people evaluated past results in the Setiqueues and pretty much concluded that the most distributed AR WU was the "0.417". I know that for myself, having had a queue record of about 1100 results at the time, I found that ~13% of the WUs were 0.417. This was out of an ENORMOUS possible number of WU ARs (considering ARs have been distributed from a range of 0.000 - ~18.000, from what people have reported).

Thus the bench was changed to a 0.417.

Anyone interested in this fascinating thread may check it out here.

[EDIT: BTW, I know that not every AR is apparently distributed as a WU, so the actual number of ARs that have been sent out is not the theoretical.]