Welcome, Guest. Please login or register.

Author Topic: Seti@home  (Read 9089 times)

Description:

0 Members and 1 Guest are viewing this topic.

Offline X-ray

  • Hero Member
  • *****
  • Join Date: Jul 2004
  • Posts: 4370
    • Show all replies
Re: Seti@home
« on: May 17, 2006, 06:47:29 PM »
That's strange, with my old mobo and OS I couldn't get the BOINC client to work, but since upgrading my hardware I seem to have no problems.
Now if only they can iron out the computing errors with the Enhanced Work Units, everything will be sweet. A computing error after 6 hours kind of sucks. On my machine the enhanced units are taking a fair amount of time to compute (about 6 to 7 hours), so I reckon on a modest system like the Sempron I had before, these times must be heading to the 12 hour mark at least  :-o
 

Offline X-ray

  • Hero Member
  • *****
  • Join Date: Jul 2004
  • Posts: 4370
    • Show all replies
Re: Seti@home
« Reply #1 on: May 17, 2006, 10:09:22 PM »
@ Hyperspeed

"...I have a feeling we'll find out in 10yrs time that what you lot were computing was not search algorithms for extra terrestrial communication attempts..."

----------------------------------------------------------------

You're joking, right?
 

Offline X-ray

  • Hero Member
  • *****
  • Join Date: Jul 2004
  • Posts: 4370
    • Show all replies
Re: Seti@home
« Reply #2 on: May 17, 2006, 10:38:45 PM »
@ Vincent

Yup, its good to put this new PC to some use now.

When you say 3 hours are you talking about the 4.18 standard units or the 5.12 enhanced units? Because I do the 4.18 units in about 2.5 hours but the enhanced 5.12 ones are definitely more than 4 hours on my setup. This is a dual core AMD 4200, so essentially it is like two CPUs at 2,2gHz. I would have thought that beats the Sempron. What units are you processing? I am being given a mix of 4.18s and 5.12s.

Edit: each core is assigned a separate work unit, and it does them simultaneously. Sometimes one core will be doing a 4.18 and the other will do a 5.12. If I display the graphics for both, I get two 3D graphs and the 5.12 will say 'Enhanced' whereas the 4.18 is the same as it always was.
 

Offline X-ray

  • Hero Member
  • *****
  • Join Date: Jul 2004
  • Posts: 4370
    • Show all replies
Re: Seti@home
« Reply #3 on: May 17, 2006, 10:51:53 PM »
Okay scratch the enhanced questions, I checked your results and you are doing the mix too. You are averaging the same processing time for the standard units as me (about 7,000 to 8,000 sec) but I have a slight edge on you with the enhanced units: my range is 16,000 to 33,000 sec with most of them being at the 21,000 mark whereas yours is more like 19,000 to 35,000. It seems to me that your Sempron is not much slower than one core of my CPU.

My higher average credit comes down to a 2 cores vs 1 performance difference (unless I have calculated wrong)

Edit: just saw that some of those results of yours for standard units were 14,000 so the Sempron is definitely slower.
 

Offline X-ray

  • Hero Member
  • *****
  • Join Date: Jul 2004
  • Posts: 4370
    • Show all replies
Re: Seti@home
« Reply #4 on: May 18, 2006, 11:29:10 PM »
@ Hyperspeed

"...This talk of the program running on two seperate cores of the CPU at the same time is plain devious in my opinion..."
----------------------------------------------------------

I suggest you look up what dual core really means and how it is supported by the various OSes in current use.

When I raytrace with Cinema4D it uses both cores much the same as Seti does: one core raytraces the top half of the image and the other one raytraces the bottom half. The only difference is that Cinema is working on separate parts of one 'object' while Seti is using two objects. The user can assign what percentage of 'resources' Seti uses. I can choose how many cores it uses, and what percentage of that core it uses. There is nothing underhand about it. So where is the chicanery that you seem so anxious to bring to our attention?
Futhermore do you really think that the clandestine use of millions of computers would go undetected when the data sets being analysed are downloaded onto the host computers and are freely available to analyse? If this was all a big conspiracy then how come third party applications designed to crunch these sets in a more optimised way are being used with no reproach?
 

Offline X-ray

  • Hero Member
  • *****
  • Join Date: Jul 2004
  • Posts: 4370
    • Show all replies
Re: Seti@home
« Reply #5 on: May 25, 2006, 06:04:19 PM »
@ Blobzie and Vincent

Now that we have these 5.15 units, there seem to be more errors, even more than 5.12. I just stopped one unit that had been going for 12 hours and was stuck on 98%. On a single core machine, that will go forever and you'll get no progress with Seti if you don't stop it.

Just thought I'd remind you to check more often on the progress of your units  :-)
 

Offline X-ray

  • Hero Member
  • *****
  • Join Date: Jul 2004
  • Posts: 4370
    • Show all replies
Re: Seti@home
« Reply #6 on: May 25, 2006, 09:20:50 PM »
Are you doing any 5.15s, B?
 

Offline X-ray

  • Hero Member
  • *****
  • Join Date: Jul 2004
  • Posts: 4370
    • Show all replies
Re: Seti@home
« Reply #7 on: May 26, 2006, 03:34:54 PM »
And how long do your 5.15s take compared to the 5.12s?
(Can't see which ones are 5.12s and which ones are 5.15s in your list, they both show as enhanced units)
 

Offline X-ray

  • Hero Member
  • *****
  • Join Date: Jul 2004
  • Posts: 4370
    • Show all replies
Re: Seti@home
« Reply #8 on: May 26, 2006, 09:12:49 PM »
@ Hyperspeed

You seem to have an affliction that reduces your ability to contribute anything worthwhile to this thread.
In the beginning it was of mild amusement, similar to the discovery of a multicoloured bird dropping on the pavement.
But now it is taking on all the charm of a donkey repeatedly braying at his own reflection in the pond.

If you want to be the source of a few giggles, I can quickly arrange it...I'll slap an ode on your arse  ;-)