Greetings fellow earthlings...

Just received my first credits Granted 38.72
Pending 266....reporting from China 6-25-04
04:00 local time

Each WU has to be confirmed with 2 out of 3 results matching. So YOU may have sent your results in, but the other people with that WU might have slower computers, so they haven't sent theirs in yet.

From what I understand, the server will wait 15 days before resending unconfirmed results, so although it COULD take 15+ days, it will probably be significantly less... 2-5 days I'd guess.

> Hm, just what I wondered !??
> I have loads of Pending credit, like 332.22 on S@H2
> but no granted credit at all !?????????
>
>
> Has anyone else actually got any granted credits yet and
> how long did they take to come in !??????!??
>
>

> Each WU has to be confirmed with 2 out of 3 results matching. So YOU may
> have sent your results in, but the other people with that WU might have slower
> computers, so they haven't sent theirs in yet.
>
> From what I understand, the server will wait 15 days before resending
> unconfirmed results, so although it COULD take 15+ days, it will probably be
> significantly less... 2-5 days I'd guess.
>
> > Hm, just what I wondered !??
> > I have loads of Pending credit, like 332.22 on S@H2
> > but no granted credit at all !?????????
> >
> >
> > Has anyone else actually got any granted credits yet and
> > how long did they take to come in !??????!??
> >
> >
>
>
Actually it is supposed to be 3 matching WUs.
jm7

12798 66.17 16264 67.60 19509 66.29 23504 66.11
Pending credit: 266.17

My machine has been sitting on these WU ID's and Pending credits for 5 days,
all still at the "initial" status...my Granted was 38.72. I figure the above
mentioned results are on slower machines, or people tank up their cache...
would be nice to get my 266.17. JM....after going to "Initial" are there still
two more steps before "Granted"? I set my prefs for 2 to 3 days of work, JM
is this too high, which would make it unfair to those waiting on the same WU..
my machine avgs. around 8 hours on each WU....

> 12798 66.17 16264 67.60 19509 66.29 23504 66.11
> Pending credit: 266.17
>
> My machine has been sitting on these WU ID's and Pending credits for 5 days,
> all still at the "initial" status...my Granted was 38.72. I figure the above
> mentioned results are on slower machines, or people tank up their cache...
> would be nice to get my 266.17. JM....after going to "Initial" are there
> still
> two more steps before "Granted"? I set my prefs for 2 to 3 days of work, JM
> is this too high, which would make it unfair to those waiting on the same
> WU..
> my machine avgs. around 8 hours or each WU....
>
What your cache settings need to be depend upon your situation. If for example, you have a dialup account that you connect through once a week, 7 to 7 days would be nessecary. If you have an on all the time connection, and are willing to work on more than one project, then a very small queue might be what is wanted. If you only want to to S@H, then two or three days of cache seems reasonable. I personally have my cache set to .1 to .2 days, and I am attached to 4 projects: BOINC Alpha (invitation only), BOINC Beta, S@H, and Predictor. Having several projects will also slow down the reporting of results to each of the projects as the results are only reported when a particular project is connected to request more work, or one of the results is in danger of expiring.

> Actually it is supposed to be 3 matching WUs.
> jm7

John, can you ask or tell what's wrong with this unit?
http://setiboinc.ssl.berkeley.edu/sah/workunit.php?wuid=17177

If it's the discrepancy of me asking so much, all I can say is that I am always asking so much. Don't want to keep the others from getting credit. ;)


----------------------
Jordâ„¢
[url=http://www.boinc.dk/index.php?page=user_statistics&userid=41965]

Crystallize, I doubt we can blame that. The one with the 3.18 result is using a Power Macintosh.

Only the Windows core client is 3.19 ..

----------------------
Jordâ„¢
[url=http://www.boinc.dk/index.php?page=user_statistics&userid=41965]

Jorden -
You may want to re-run your benchmarks, I notice that your computer seems to be claiming pretty high credits on most of it's completed WU's, and the benchmark figures seem pretty high for your processor type.

> Jorden -
> You may want to re-run your benchmarks, I notice that your computer seems to
> be claiming pretty high credits on most of it's completed WU's, and the
> benchmark figures seem pretty high for your processor type.

Been there, done that. Just done it again even.
But even throughout Beta my CPU was asking higher credits than an equivalent full P4. Maybe it's BOINC not knowing what to do with a P4 Celeron 2.3GHz that's overclocked to 2.76GHz. ;)

But okay, benchmarks according to site prior to benchmarking:
Measured floating point speed 1775.04 million ops/sec
Measured integer speed 3752.22 million ops/sec
Measured memory bandwidth 953.67 MB/sec

Benchmarks just done in the GUI with everything else that was running off:
--- - 2004-06-25 15:22:14 - Benchmark results:
--- - 2004-06-25 15:22:14 - Number of CPUs: 1
--- - 2004-06-25 15:22:14 - 1775 double precision MIPS (Whetstone) per CPU
--- - 2004-06-25 15:22:14 - 3752 integer MIPS (Dhrystone) per CPU
--- - 2004-06-25 15:22:14 - Finished CPU benchmarks

Running again.
--- - 2004-06-25 15:27:07 - Benchmark results:
--- - 2004-06-25 15:27:07 - Number of CPUs: 1
--- - 2004-06-25 15:27:07 - 1781 double precision MIPS (Whetstone) per CPU
--- - 2004-06-25 15:27:07 - 3873 integer MIPS (Dhrystone) per CPU
--- - 2004-06-25 15:27:07 - Finished CPU benchmarks

And again.
--- - 2004-06-25 15:28:44 - Benchmark results:
--- - 2004-06-25 15:28:44 - Number of CPUs: 1
--- - 2004-06-25 15:28:44 - 1776 double precision MIPS (Whetstone) per CPU
--- - 2004-06-25 15:28:44 - 3832 integer MIPS (Dhrystone) per CPU
--- - 2004-06-25 15:28:44 - Finished CPU benchmarks

It doesn't matter much.


----------------------
Jordâ„¢
[url=http://www.boinc.dk/index.php?page=user_statistics&userid=41965]

> Thanks for the info, it doesn't seem very promising, +15 days.
> Now I have pengding 439.47 granted 0.00
I have over 10,000 pending and only 28.35 granted!
I participated in the Beta and understand this will take time, you guys have NOTHING to complain about!
I have 7 computers crunching Boinc now and 10 others still crunching on Seti Classic.
Calm down and let the units start coming in, we are still in the intiial stages of the program.

Pending credit: 592.92 Total credit 39.17 at this time
and------------
---------------------------------------------------------------------------
--- - 2004-06-25 15:50:57 - Cache low-water mark hit; requesting more work
SETI@home - 2004-06-25 15:51:01 - Message from server: Not sending work - last RPC too recent: 27 sec
SETI@home - 2004-06-25 15:51:01 - Deferring communication with project for 57 minutes and 3 seconds
SETI@home - 2004-06-25 15:51:01 - No work from project
------------------------------------------------------------
shit happens ;-)
[/url] [/url]

IN MEMORY OF CARL SAGAN.1934-1996

Guido can you please refrain from having so many different and large pictures as a consideration to those of us on 56k please?

OOPS i hope this is better ;-)
[/url] [/url]
IN MEMORY OF CARL SAGAN.1934-1996

> > Jorden -
> > You may want to re-run your benchmarks, I notice that your computer seems
> to
> > be claiming pretty high credits on most of it's completed WU's, and the
> > benchmark figures seem pretty high for your processor type.
>
> Been there, done that. Just done it again even.
> But even throughout Beta my CPU was asking higher credits than an equivalent
> full P4. Maybe it's BOINC not knowing what to do with a P4 Celeron 2.3GHz
> that's overclocked to 2.76GHz. ;)
>
> But okay, benchmarks according to site prior to benchmarking:
> Measured floating point speed 1775.04 million ops/sec
> Measured integer speed 3752.22 million ops/sec
> Measured memory bandwidth 953.67 MB/sec
>
[snip]
>
> ----------------------
> Jordâ„¢
> <a> href="http://www.boinc.dk/index.php?page=user_statistics&userid=41965"><img> height="70"
> src="http://www.boinc.dk/auto.php?user=41965&project=sah&input=1071259300+-+earth2&layout=1071259300+-+earth2.jpg">
>
I do not trust these benchmarks very much.

In a former life I worked on optimizers for C compilers. The Whetstone benchmark was widely believed to be a test of floating point computation, but the module that actually contained floating point code is in there to test call and return efficiency, not floating point performance. And a good optimizer, of 1990 or a bit before vintage, expands that module in line so it does not test call and return efficiency at all. And once expanded in line, the loop-invariant-code motion moves all the floating point operations out of the loop (that goes around 10000x, IIRC), so they are done only once. Then the live-dead analysis removes all of them because the results are not used. So on a modern compiler, it is by no means a test of floating point performance.

Now I have two machines that run BOINC, and they are extremely different. One is a 1996 era machine with a Pentium 166 MHz CPU. It was not state of the art when I bought it. And this machine I just built and was nearly state of the *86 art at the time I started purchasing the parts (but not by the time I finished it). Uses Intel e7501 chipset. Yet their memory bandwidth comes out the same. This aroused my curiousity and I looked at the memory bandwidth of other machines that are posting BOINC results and they all have the same memory bandwidth.

It says I have 512K Cache, but I have two 1024K caches, one on each CPU. Perhaps they divide by the apparent number (4) of CPUs, not the actual number (2 hyperthreaded ones).

So I would not take these benchmarks too seriously.

> Yet their memory bandwidth comes out
> the same. This aroused my curiousity and I looked at the memory bandwidth of
> other machines that are posting BOINC results and they all have the same
> memory bandwidth.
>
> It says I have 512K Cache, but I have two 1024K caches, one on each CPU.
> Perhaps they divide by the apparent number (4) of CPUs, not the actual number
> (2 hyperthreaded ones).
>
> So I would not take these benchmarks too seriously.

Regardless of the other stuff, I do recall a mention that the memory bandwidth was no longer being calculated, and they just stuck a fixed value into the database.

Hi

I have 3 machines running, me, my brother and my son.
We all have an AMD Athlon XP 2800+ and all benchmarks are the same between 20 points.

greetz Mike

> whats with the double post's?

The below post is soley aimed at Crystallize-LKSAB

> I also participated in the Beta test since Mars 2003,
> but I have the right to complain all I want.
> I think S@H 2 s*cks
> If it wouldn't for that I really believe in the research I would quit
> and turn off my computers at night.
> I think its sad that they have "destroyed" such a nice project with
> this kind of gibberish credit system.

Since you participated in the Beta, you'd be better of out there in the Q&P helping the completely new people to get their stuff running instead of complaining here.

I have never seen you complain about this on the Beta forums, so don't start here and be a menace. We're only waiting for JQ to find these forums and be a mess besides you, if you want.

So if you ever thought it S*cked, why not give a descript definition of how, where and what sucked about it to the developers so that they could make it different?

Don't go complaining on here if you did Beta. It doesn't suit.
You're an arsehole if you all of a sudden "think you see the light" and try to get all the other SAH Classic people behind you.


----------------------
Jordâ„¢
[url=http://www.boinc.dk/index.php?page=user_statistics&userid=41965]