• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Insider Seti@home info?!

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.


<< Escalade

It would be very helpful if you said where that discussion came from
>>




OOPS! I mean to do that!

It was in both the sci.astro.seti and alt.sci.seti usenet newsgroups the early part of last week.

I don't know how factual either satement is, but...

We all know about the problems the seti@home servers have had over the past month do to just a 30% increase in usage. That coupled with the fact that they now have ~50% more computer power than workunits - it's not very effective to have the same workunit processed more than once. Lowering the amount of processor ultization (or additional science[?] with some 'busy work' processing - if it was important science why wasn't it being processed before?) would solve both of these problems for them - i.e. The way I see it is they operate on a shoetring budget, and this is a very simple, low cost solution to their 'problems.'


 
Assimilator1,

I don't think I have any program like that.
I have win98SE. Anything like that in it?
Which one I should get?

SETI has been running all the time on this machine(Athlon 700) and in less than 24 hrs I have gotten 75 WU's in RC5.

EDIT: I just checked, It was 10.5hrs for the last 32 WU's.

I have to get going now. I will try the SETI priority later.
 
Escalade - in order for this to at least appear to be a valid scientific project, the WU results must be run multiple times to verify the data - especially since it's being run using so many different clients on so many different machines and platforms.

However IMHO, the use of &quot;busy work&quot; to alleviate a network problem sortof puts the &quot;science&quot; second and is only a bandaid. There are other things that could have been done immediately to help conserve the limited bandwidth until a more permanent solution was found.
 
Poof, just saw your posts, as a guess the pattern that alastair b found is likely introduced by whatever FFT is used by the S@H team. If you are familiar with filter design that type of pattern is common...
 
HB - it's still cool as hell though! 😀

Curt - if you're running 9x, you may have to get another proggy to do it for you, eg., Seti Driver lets you do it - and maybe something like Taskinfo too? (haven't used that program yet). With NT/2K, you can do it through task manager.
 
actually, if you look at the results received vs wu's generated on the s@h website, they've run each unit 3.4 times.
 
Curt Oien

To check cpu utilisation (&amp; alter SETI's priority??) use this program

Esclade
Thanks for replying ,(aside from my 1st post😛) ,it seems the verdict is still out as to why WU times are much longer (aside from the extra science which could be added because of all the extra power at hand).
I still doubt that they would deliberatly reduce the cpu useage though.
 



<< I still doubt that they would deliberatly reduce the cpu useage though. >>




I wouldn't be so sure of that...

Remember the patch that was written by Olli that replace the FFT routine - made Seti@Home run quite a bit faster.

Also Microsoft optimized Seti@Home for Intel-III processors/Windows. Once again Seti@Home was shown that it could run faster.


But what did the Seti@Home guys do... did they look a what they were doing and try to improve upond it... NO, they balked at it! and what was their reason... their algorithms might return incorrect results! This simply isn't true, and if they were concerned they could easily test it, 2 plus 2 is going to equal 4 in each and every case. The (sad) fact of the matter is they simply can't support ($) users crunching workunits any faster.

If this was truly a scientific search, they would optimize the program for each of the major processor/operating enviroments so results can be processed as fast as possible, but they simply can't meet that kind of supply with their current setup.


Sorry for the negative tone, but it irks me to no end, that all the excess processing power they have available to them is going to waste doing 'busy work.'


 
Its 1 thing choosing not to use an optimised client (like you said they wouldn't be able to handle the extra load) ,but its a another thing to ramp down cpu utilisation ,I still doubt they would do that !😛😉

but they simply can't meet that kind of supply with their current setup.
I agree ,but at least we have the project.

Sorry for the negative tone, but it irks me to no end, that all the excess processing power they have available to them is going to waste doing 'busy work.'

Thats not proved yet..........
 
Assimilator1,

Thanks for the task manager.

SETI(3.03 GUI)runs at about 80%
RC5 is at about 14%
and the task manager itself uses about 4%
SETI and RC5 both show a priority class of idle (4)

If the task manager was not running that would give 80% to SETI and 18% to RC5. That is close to the 20% I had guessed at.

When I shut down RC5, SETI jumps to 96%.
So much for RC5 getting out of the way for everything else.
 
No probs 🙂 ,thanks for the info BTW.
Ref RC5 no longer sitting idle with SETI,maybe there are different degrees of idle priorites??? ,anyone?
Does it make any difference which 1 is started 1st?
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top