Hey Orange Kid

[TuffGuy]

why's my efficiency so horrible? it shows 95% efficiency on my home setiq and 13% on yours.

 

[Russ]

Actually, I've been wondering just what those numbers mean.

Russ, NCNE

 

[TuffGuy]

methinks that it's a measure of the time that you have seti cracking. for me, i have my computer running seti ~95% of the time.

 

[Russ]

HMM...I don't think that's it. All of mine crack 24/7 and I have some that are way down in the teens.

Russ, NCNE

 

[TuffGuy]

yeah, that's why i'm trying to reach OK. there's something wrong...

 

[TuffGuy]

i mean, i've downloaded 351WUs from his proxy and uploaded 633WUs. that's *pretty* efficient.

 

[Eponymous]

I think what happens is this:

Your Queue talks to his Queue.

His Queue thinks you are just one machine and takes your total time
and tries to figure out how efficient your one machine is for the
number of packets it is handing out.

I have noticed that I store some packets local on each machine.
Say 4 WUs, because it takes a while to go through all four WUs.
I always have 25% efficiency.

Does that make any sense?

 

[Smoke]

I have no idea what that means either but I just wanted to add that http://www.smokeball.dns2go.com:5517/userinfo.htm?user=2978765 shows 100% Efficiency.

That's a laugh. My "Garbage Brigade" is 100% Efficient! LOL

 

[TuffGuy]

a little... kind of a backwards way of keeping track of each machine but...

 

[TuffGuy]

lol @ smokeball

 

[IBhacknU]

I think Eponymous is on track. SetiQ doesn't really take into account programs like SetiDriver that 'stock up' on WU's

take this example:

Average Time 2 days 19 hr 28 min
Average CPU Time 8 hr 31 min (12% efficient)

This is a machine running 24/7 with a cache of 8 WU. By the time a WU is crunched and returned, SetiQ has had to wait for 7 other WU's to process.

8 x 8.5 hrs = 2 days, 19.5 hours (roughly)

Well, the more I think about it, the more I think efficiency is not a good way to describe it. I mean, If I'm running Photoshop or some Video edititng stuff, the WU's take much longer to process. I can end up with a 10 hour WU (using the above example).

CPU time only represents how long the PC took to process the WU, not what else the processor was doing.

It's too late and this really doesn't matter (to me). The fleet is on cruise control, and so am I.

out!

 

[JWMiddleton]

Yep, Eponymous and Hack are correct. The only way to have efficiency show a meaningful number is to get 1 WU at a time. Not worth it to me!

 

[Smoke]

 

[Orange Kid]

Well i'm late to this thread, but what they said

The times on a seti wu are essentially meaningless......all that really matters to Berkeley are the results. Whether it be a "5 minute" wu or a "5 day" wu as long as the result is duplicated by someone else, then that is good data

It is all about the science after all isn't it?

Sorry if this reply sounds a little cranky, I am slightly dejected by the way things are going around here right now

 

[Orange Kid]

All that and I didn't really answer the question

SetiQ has your average time per wu as 3hrs+ and you connect only once a day as an average.....so 3 divided by 24 gives you a 12.5% efficiency ratio. SetiQ thinks you are doing too many other things with your computer...../me shrug his shoulders....one of those not so good stats that the Q generates