> Runs fine on my Pentium 75 with 72mb RAM. :)
>

Hmm, I think the only way I can beat that would be to mosix cluster my Pentium 100 with 64mb ram to my Pentium 75 that i'd throw 32megs of ram into it, then run two boinc's on the Pentium 100 and one of the seti processes would migrate to the Pentium 75 :)

Well I guess I could always underclock the P100 but I don't think thats fair.

---

I do have a 486 dx2 66 with 24 meg ram i could copy a boinc install to and see if i can force it to run, see how long it takes, I of course would make it do a WU already downloaded by one of my other machines but not started or barely started on as i dont want to make everyone waiting for credit to have to wait for it to expire.

The question is should I?

---

I think you will need at least 64 MB RAM free

---

Anyone here have a 386sx 16 they can underclock to 12 or 10, it can be done by changing out the rectangular 4 pin osolators, and 4 16 meg 30 pin simms, they do exist, and run boinc on that :P

---

Actually, apparently the 128mb requirement is artificial,
prior to BOINC 4.19 SETI@home ran happily in 64mb of RAM.

There is a thread discussing this at
Message Boards:Number Crunching:64 MB enough memory?

---

Need help? Check out the excellent Unofficial BOINC-Wiki!
'We are the BOINC. Prepare to be assimilated.'

> Actually, apparently the 128mb requirement is artificial,
> prior to BOINC 4.19 SETI@home ran happily in 64mb of RAM.

I just checked one of my machines running Boinc 4.21 and it's memory useage is peaking at 3.2% of 512Mb. ;)

---

Be lucky

Neil

> > What happens if user interest is so high that there is a greater demand
> than
> > one million results a day? Maybe we'll generate more, but remember the
> real
> > beauty of BOINC (and one of the reasons it was created) - you can crunch
> on
> > other projects if any particular project is temporarily out of work.
> >
> > - Matt
> >
> Might I be so bold as to suggest that the new users be limited in their cache
> size until they understand the way the system works. With say 100,000 users
> each asking for 10 days of work on each of who knows how many computers, the
> units that need to be available could be a whole lot more than a million.
> I know that now users are limnited to a max of 10 days per computer, but some
> of use have farms..if we all cranked up that cache to the full 10 days worth
> you would be out of units in no time!
> Maybe even lower us "always on" people to fewer days than the dial-up users.
> You should be able to figure out from how often the units are returned by each
> of us, some guidelines.
>
>Mikey

take a look at my message "Connect to network about every" or "No Work Available" 77734 and tell me what you think

---

> > according to setisynergy.com the total numbers of Boinc users is 72,589
> as of
> > 19 hours ago.
> >
> So, if my guess is right, then 4x72,589=290,356 users is 100% load. That
> means the server can accept up to 217,767 new/classic transferred users.
>
>
>

Well, don't forget that the new machine has room for 2 more CPUs, and 4 times as much RAM.

---

> > > according to setisynergy.com the total numbers of Boinc users is
> 72,589
> > as of
> > > 19 hours ago.
> > >
> > So, if my guess is right, then 4x72,589=290,356 users is 100% load.
> That
> > means the server can accept up to 217,767 new/classic transferred users.
> >
> UMMMM...not exactly, but close. Remember the guy who started this whole thread
> said he had created a Boinc account but never used it. There are probably more
> like him.
>
>

I just checked at www.boincstats.com the users with more than 0 credit for the last month = 38,357

That equates to approx. 1/2 of the accounts NOT used in the last 30 days.

---

> Well, don't forget that the new machine has room for 2 more CPUs, and 4 times
> as much RAM.
>
Excellent point.. I had forgotten about that.

thanks

---

> >
>
> Well, don't forget that the new machine has room for 2 more CPUs, and 4 times
> as much RAM.
>

Doh!

---

You will be assimilated...bunghole!

> > > > according to setisynergy.com the total numbers of Boinc users
> is
> > 72,589
> > > as of
> > > > 19 hours ago.
> > > >
> > > So, if my guess is right, then 4x72,589=290,356 users is 100% load.
>
> > That
> > > means the server can accept up to 217,767 new/classic transferred
> users.
> > >
> > UMMMM...not exactly, but close. Remember the guy who started this whole
> thread
> > said he had created a Boinc account but never used it. There are probably
> more
> > like him.
> >
> >
>
> I just checked at www.boincstats.com the users with more than 0 credit for the
> last month = 38,357
>
> That equates to approx. 1/2 of the accounts NOT used in the last 30 days.
>

Hello
All of you discuss current and maximum expected number of users (active or nor). In my opinion we should take a look at number of active hosts.

---

BOINC@Poland: Polish crossproject team.
http://mchl.republika.pl/boincatpoland/
- Join -

> > > > > according to setisynergy.com the total numbers of Boinc
> users
> > is
> > > 72,589
> > > > as of
> > > > > 19 hours ago.
> > > > >
> > > > So, if my guess is right, then 4x72,589=290,356 users is 100%
> load.
> >
> > > That
> > > > means the server can accept up to 217,767 new/classic
> transferred
> > users.
> > > >
> > > UMMMM...not exactly, but close. Remember the guy who started this
> whole
> > thread
> > > said he had created a Boinc account but never used it. There are
> probably
> > more
> > > like him.
> > >
> > >
> >
> > I just checked at www.boincstats.com the users with more than 0 credit
> for the
> > last month = 38,357
> >
> > That equates to approx. 1/2 of the accounts NOT used in the last 30
> days.
> >
>
> Hello
> All of you discuss current and maximum expected number of users (active or
> nor). In my opinion we should take a look at number of active hosts.
>
Yeah, Sometimes users start sharing an account to make a larger credited account and such, host number means more than users.

---

Anyone here have a 386sx 16 they can underclock to 12 or 10, it can be done by changing out the rectangular 4 pin osolators, and 4 16 meg 30 pin simms, they do exist, and run boinc on that
I've got a IIcx (68030@16 8MB/1.2GB) and a Performa 6360 (603ev@120 136MB/1.2GB) that I've wanted to get crunching, but NuBus ethernet adapters are even more expensive than they were five years ago, and I've had trouble compiling the Linux driver for a D-Link DFE-530TX+ NIC under Yellow Dog 3.0.1.

---

> Hello
> All of you discuss current and maximum expected number of users (active or
> nor). In my opinion we should take a look at number of active hosts.
>

Good Point!

There is a total of 178,213 hosts.
And only 68,985 hosts have produced credit in the last 30 days.

Approx, only 1/2 the users have produced credit in the last 30 days.

That's 38.7% of the total number of hosts that have produced credit in the last 30 days.
And 52.5% of the users that have produced credit in the last 30 days.

So, that is lot of hosts & users that are NOT taxing the serving.
Well, I presume so, since they're obviously not contacting the server to return results.

---

> > Hello
> > All of you discuss current and maximum expected number of users (active
> or
> > nor). In my opinion we should take a look at number of active hosts.
> >
>
> Good Point!
>
> There is a total of 178,213 hosts.
> And only 68,985 hosts have produced credit in the last 30 days.
>
> Approx, only 1/2 the users have produced credit in the last 30 days.
>
> That's 38.7% of the total number of hosts that have produced credit in the
> last 30 days.
> And 52.5% of the users that have produced credit in the last 30 days.
>
> So, that is lot of hosts & users that are NOT taxing the serving.
> Well, I presume so, since they're obviously not contacting the server to
> return results.
>

Actually, the number of inactive users/hosts does 'tax' the server because it makes the database table that much larger. Thus, when the active users/hosts contact the server, the table takes longer to look through.

I expect some of the classic users will not go to BOINC. So, if there are about 450,000 active classic users, maybe 80%? of those will convert. I think that many crunchers are college/university students and quit SETI@Home when they graduate. If the transition to BOINC happens in the late spring this year, probably many of those students will not bother to sign up for BOINC just for a few weeks.

---

> Actually, the number of inactive users/hosts does 'tax' the server because it
> makes the database table that much larger. Thus, when the active users/hosts
> contact the server, the table takes longer to look through.

It does, but the size of the table is not that significant. Especially when you consider that the entire database of SETI@Home users was migrated at one point last year and we have been working against that full set of users since then.

So, all that would be added to BOINC are the users that have signed up for Classic since the snapshot was taken in March 2004 ...

---

> So, all that would be added to BOINC are the users that have signed up for
> Classic since the snapshot was taken in March 2004 ...
>
I suspect that what would be added 'for handling purposes' would be not only the additional users to Seti Classic since last March (a fairly large number perhaps around a quarter of a million), but also a very real surge of Seti Classic users who at least try the BOINC model (lacking an alternative), plus that portion of Classic users who try and stay.

Barry Schnur

---

> > Actually, the number of inactive users/hosts does 'tax' the server
> because it
> > makes the database table that much larger. Thus, when the active
> users/hosts
> > contact the server, the table takes longer to look through.
>
> It does, but the size of the table is not that significant. Especially when
> you consider that the entire database of SETI@Home users was migrated at one
> point last year and we have been working against that full set of users since
> then.
>
> So, all that would be added to BOINC are the users that have signed up for
> Classic since the snapshot was taken in March 2004 ...
>
>

It is true that the users have been migrated, but unless they become active, their is no host info.