After updating the nbody validator it's now using boinc's default credit system. I'm wondering how this credit compares to what was being awarded before, and if we'll need to apply some kind of modifier to it or not. So let me know how it's doing.

---

It is doing well. It seems better balanced now.

Yes, of course it still favours modern CPUs with lots of cores and BIG L3 cache and fast memory. But not so much like before.

Its because of the nature of the NBody app which is pretty data-intensive, using 25 MB in RAM (both versions, 64-bit and 32-bit). This means that memory speed and L2 + L3 cache helps a lot.

--
My data for Intel Core i7 720QM, 8 logical cores at 1.73 GHz (Clarksfield, 45 nm, TDP 45 Watt), 1MB L2 + 6MB L3 cache, 4x2GB DDR3 1333 CL7 in DUAL ch.!
--------------------------------
OLD NBody units had this ratio:
.... 204.25 (sec. of CPU time) .......... 3.00 credits.
NEW NBody units computed today have this ratio:
.... 553.27 (sec. of CPU time) .......... 2.98 credits.
--------------------------------

For Intel Atom N270 at 1.6 GHz, 512MB L2, (2 logical cores withOUT out-of-order execution[b] !, a feature in all other CPUs making them run much faster in non-optimized code), DDR2 533 RAM.
--------------------------------
OLD NBody units had this ratio:
.... [b]1,142.89 (sec. of CPU time) .......... 3.55 credits.
NEW NBody units computed today have this ratio:
.... 1,186.56 (sec. of CPU time) .......... 3.57 credits.
--
You see? The EXTREMELY TOO-FAST CPU suffered some penalty. The veeeery slow CPU did not.
More balanced. :-)

I need comeone to confirm the Intel Atom change in credits, because my "NEW" WU has started yesterday!!

It just finished (only last 33 percent of it) today at 9:00 CET.

I suspect it may have been rewarded by OLD credit system!
So I am sorry if the data for Intel Atom are wrong. So, be warned - and do NOT rely on them. :)

ALERT! ALERT! ALERT!

Still NO WORK available from project Milkyway at Home !

Please, create some. :)

Fr 01 Jun 2012 14:43:34 CEST | Milkyway@Home | update requested by user
Fr 01 Jun 2012 14:43:36 CEST | Milkyway@Home | Sending scheduler request: Requested by user.
Fr 01 Jun 2012 14:43:36 CEST | Milkyway@Home | Requesting new tasks for CPU
Fr 01 Jun 2012 14:43:38 CEST | Milkyway@Home | Scheduler request completed: got 0 new tasks
Fr 01 Jun 2012 14:43:38 CEST | Milkyway@Home | No work available


Going in the town for shopping.....

second day now...
6/1/2012 11:26:13 AM | Milkyway@Home | Requesting new tasks for ATI
6/1/2012 11:26:15 AM | Milkyway@Home | Scheduler request completed: got 0 new tasks
6/1/2012 11:26:15 AM | Milkyway@Home | No work available

Is it just me or is there really no work ? Should I switch to something else for a while ?

second day now...
6/1/2012 11:26:13 AM | Milkyway@Home | Requesting new tasks for ATI
6/1/2012 11:26:15 AM | Milkyway@Home | Scheduler request completed: got 0 new tasks
6/1/2012 11:26:15 AM | Milkyway@Home | No work available

Is it just me or is there really no work ? Should I switch to something else for a while ?

Not just you, same here. I'm runing POEM@home while Milkyway is low on GPU wu's.

I'm finishing some kinks on the new work generators (unfortunately had to sleep in the middle of it). Work should be generating in an hour or two.

---

All I get for nbody since 1. june is:
04.06.2012 21:15:24 update requested by user
04.06.2012 21:15:29 Sending scheduler request: Requested by user.
04.06.2012 21:15:29 Requesting new tasks for CPU
04.06.2012 21:15:30 Scheduler request completed: got 0 new tasks
04.06.2012 21:15:30 No work available

Popsie--might want to check the Number Crunching forum...we are aware of the no work issue and it has been escalated:

http://milkyway.cs.rpi.edu/milkyway/forum_thread.php?id=2917

---

Popsie--might want to check the Number Crunching forum...we are aware of the no work issue and it has been escalated:

http://milkyway.cs.rpi.edu/milkyway/forum_thread.php?id=2917

It's a power outage issue, see the post I just made. Hopefully I'll be able to fix things soon, but right now my hands are tied since I can't log into the server.

---