LHC@home

[QUOTE from the Frontpage]

When we re-open for production running in a while, we will announce it and then welcome a bigger user base.

[/QUOTE]

a big LOL about this!

@the LHC-team;
you don't arrive to get enough work for "20 concurrent connections" , and now you announce that you will increase the user number!!

That is also a kind of deadpan humor ( or SOHF : (sense of humour failure))! LOL ;-)

-no offense, just to point how realistic this project is.-

greetz from Switzerland
littleBouncer

---

yeah, I agree... the frontpage posting really doesn't make me happy.... they better come up with more work... and I mean a LOT more!!!!

sysfried

---

> yeah, I agree... the frontpage posting really doesn't make me happy.... they
> better come up with more work... and I mean a LOT more!!!!
>
> sysfried
==========

Sigh, or what sysfried, I don't think any of us are in a position to demand anything from any of the Projects. Nobodys forcing anybody to run this Project or any of the other Projects either, if people don't like the way things are going here then run another project more to their liking ...

Hey, come on people, remind we had in the past weeks 100's of WU's with 00:00:00 CPU time.

They were throw away in a minimum of time.

---

Best greetings from Belgium
Thierry

> Hey, come on people, remind we had in the past weeks 100's of WU's with
> 00:00:00 CPU time.
>
thats true. Also true we had some returning much credits.

What makes me sad is, without that, 25% (a guess) of the clients could be turned off to produce the same amount of propper credited work

All of you are right..

*my* interpretation/understanding of the "news" mentioned by littleBouncer
is:

-they will NOT open (now) for more public
(good idead!)

-they are in "beta" (thats probaly we all *g*)

-they talk about the "bugs" and are concerned about..
(--> hope!)

>When we re-open for production running in a while
this can be a source of misunderstanding.
production reopening for us (6000 Semi Beta Testers) or
reopening for everyone?
And in which time frame?

still waiting patiently
(but it's hard)

friendly
ric

Hi Poorboy.

From my previous postings you should know that I'm not someone demanding anything from LHC. I'm always with LHC when LHC has work for my machines. Other than that, I do have other things for my CPU's to do in their idle time.

I remember a posting from LHC Admins saying that they're overwhelmed with all the cpu power they're provided with and hinting that other projects running at CERN might jump the gun.

Reading my old posting again, I see the line with the "or else".. sorry about that, I didn't mean it that way. I think small steps first.... They upgraded the servers and now can handle more clients. Now they're fixing new bugs.... when all bugs are gone (or let's say the big ones), CERN will go public... And propably provide us with more work. We'll see how much work they have for us. :-)

CHeers,

Sysfried

> > yeah, I agree... the frontpage posting really doesn't make me happy....
> they
> > better come up with more work... and I mean a LOT more!!!!
> >
> > sysfried
> ==========
>
> Sigh, or what sysfried, I don't think any of us are in a position to demand
> anything from any of the Projects. Nobodys forcing anybody to run this Project
> or any of the other Projects either, if people don't like the way things are
> going here then run another project more to their liking ...
>

---

So far we empty any queued wu's very quickly....I wonder how many wu's they expect to make once they go public?

---

Paul
(S@H1 8888)
BOINC/SAH BETA

> So far we empty any queued wu's very quickly....I wonder how many wu's they
> expect to make once they go public?

Just a thought here- consider this for possibility's sake: What if they wanted results *faster* and were able to distribute a given amount of work to 12,000 clients that would be returned in, say 1.2 days instead of giving the same amount of work to 6000 clients that would take 2.4 days or more to complete all work, hypothetically speaking... maybe they want to decrease turnaround time!

It's all a guess, IMHO. ;)

---

And propably provide us with more work. We'll see how much work they have for us. :-)
===========

There will probably never be enough work to satisfy everybody sysfried, thats the way it's been since the Project opened, you've been around the LHC Project long enough to know that.

The Dev's have openly stated that they will release so much work & then have to analyze it & during the analyzation period there will be no work or very little work at best. We all have to live with that fact and just run some other project while waiting for work from this one ...

Hehe!

When the project first opened the target was 50,000 in 50 days! They quickly changed that to 'many thousands'. It seems that 6000 is more than enough, and since a fair number are no longer active they're probably running on around half that.

At the rate that Intel increase the speed of their processors when LHC re-opens thay'll probably only need 3 PCs! :-))

---

Gaspode the UnDressed
http://www.littlevale.co.uk

some words of explaination:

What I read on the frontpage is a contradiction (also my writting is a contradiction , to point on the realisme (maybe this wired thinking , from my side, wasn't understood from several readers) )!

from Poorboy's statement:
>>The Dev's have openly stated that they will release so much work & then have to analyze it & during the analyzation >>period there will be no work or very little work at best.

on LHC frontpage they announced:
9.3.2005 16:30 UTC
Bug fixing and testing will continue until next week.
In the mean time there will be very little work available.

Why they can not write: "In the next 2 weeks there is no work!(or we are during analyzation)"?
No they write:" In the mean time there will be very little work available."

and 5 days later they announce they will increase the base number of users
when they will "re-open for production running in a while"!
"in a while" means in this case 'in the very near future', IMO.

at Sysfried:
I have the impression : You got it, what I wanted to say

at Ric:
[QUOTE from you]
>When we re-open for production running in a while
this can be a source of misunderstanding.
production reopening for us (6000 Semi Beta Testers) or
reopening for everyone?
And in which time frame?
[/QUOTE]

the "bolded" that are the points!

at MikeW

>>At the rate that Intel increase the speed of their processors when LHC re-opens thay'll probably only need 3 PCs! :-))

exactly!

greetz littleBouncer


[BTW]when I entered this reply, it hasn't all written as I wrote (some lines where gone, maybe some missing or wrong HTML-tags). After correction : it seems , 'the writting lacked coherence'. Sorry for that.

---

@littleBouncer

You should have named yourself littleChicken because you've got to be the most impatient person that runs the Projects, as soon as there is one little blip at any one of them your the first one to start running around screaming that "The Sky Is Falling...The Sky Is Falling" ...

You need to just take a couple of Valium's and learn to relax a bit more ... Thats what the other Projects are for, when theres a problem at one of them then just run another one ...

@PB

Do you speak from yourself?

So I got to know you, when I entered to this project!

greetz littleBouncer

I wouldn't use the same words as PoorBoy, but generally I must agree.

Whenever a Project takes a break (for whatever reasons), it seems to split the Community into two types of Users :

Type a)
Impatient, demanding, quickly dissatisfied and eventually frustrated.

Type b)
Patient, often has other Projects running parallel anyway, usually back with the Project the Minute it reopens with fresh work; just as it never went down, calm, professional business as usual.

I don't get it, all this is free, voluntary work; and I don't quite imagine LHC currently halting all of Cern's science Projects, just to get LHC@Home back to work.

They're a limited amount of people, given a limited amount of Resources and even they need to sleep once a while OR may be looking forward to a weekend OR (now get this) spend some quality time with their friends and families outside Science (how dare they *g*) ;)

Whoever is free from all those factors and (by all practical means) a 100%, never-sleeping workaholic with no life... he shall throw the first stone.

(heck, I'm pretty freakozoid already, but not even I could honestly pick up that first stone)

---

Scientific Network : 45000 MHz - 77824 MB - 1970 GB

> I don't get it, all this is free, voluntary work; and I don't quite imagine
> LHC currently halting all of Cern's science Projects, just to get LHC@Home
> back to work.

My memory tells me that there was an indication from one of the project people that the success and speed of the BOINC "Super-computer" opened eyes back there.

So much so that other places in CERN are thinking of creating Science Applications for their work.

Watching all 5 of the active projects one can see that, even though we have only a small group, each of the projects is having their work done at very high rates of speed.

I mean, we have on SETI@Home about 10% of the user base that is on SETI@Home Classic, and they have a hard time keeping us supplied with work. Predictor@Home made similar mention on the tpage as to the mass of work done over a weekend.

Of course, the only thing that gives me more of a lift are the rumors about the next top end Power Mac ...

---

Agreed, in fact each Project's Servers or backend Infrastructure always turned out to be the limiting factor to keep people busy with work, rather than the masses stalling on Computing Power ;)

But well, that's the "Community Supercomputer Effect" :D

Too bad there isn't a figure indicating the sustained, total TFlops/sec for each Project and BOINC in total.

---

Scientific Network : 45000 MHz - 77824 MB - 1970 GB

Well, there is this:

Users	             126,371         +1,205
Teams	              15,054           +104
Countries                200              0
Total Credit   1,619,856,256    +13,375,104
	
BOINCstats:
Currently in database:
Users    294,653
Hosts	 362,413
Teams	  33,147
Countries    908

From 13 Million plus Cobblestones I suppose you could calculate out what that means in FLOPS. I am not sure I am up to the challenge of the math...

1/100 of a day of 1,000 Double Precision FLOPS ...

---

> Total Credit 1,619,856,256 +13,375,104

Got it :
100 Credits per day equals a 1 GFlop/sec machine running 24h.

So in the last 24hours (based on XML exports), BOINC did a sustained total of :
133751 GFlops or 133.751 TFlops/sec

But IMHO, that sounds way too high, SETI Classic did about 70 TFlops/sec in its best times (with 600000 Active Users, using some 900000 active Hosts)

Either the granted Credits are significantly too high, or the Cobblestone System is grossly off.

---

Scientific Network : 45000 MHz - 77824 MB - 1970 GB

oops, I wrote it wrong...

It is 1/100 of a day of a 1,000 MIPS machine...

So I get 100 Cobblestones = 24,000 MIPS or 24 GFlops

> Either the granted Credits are significantly too high, or the
> Cobblestone System is grossly off.

Or both ... :)

---

Oh, I screwed up then (math in public *g*)

So make that 5.57 TFlops, that's lower than I expected but likely at least closer to the Ballpark.

I would estimate the actual number around 10-15 TFlops as of now.

---

Scientific Network : 45000 MHz - 77824 MB - 1970 GB