> I have been doing this a lot longer than you have been Aaron, and I am
> using a P90, and a P233. For someone to tell me that I CAN'T process
> first LL tests and have a chance, (slim as it may be), to find the next
> largest prime is total B.S. I am not high on the rankings, because some
> of the LL tests were before we had prime net and because I have "only"
> two "slower" machines. (They are MY machines and not some corporation's
> or university's machines, that I have loaded the software on and am
> getting credit for.)
Yikes! :-)
I've been doing this for a while too (pre-Primenet) and if I recall, at the
time I was using P-100's up to P-133's. Those were fine for exponents in
the 1M-2M range for first time LL tests. I think I began with GIMPS around
December '96 or January '97, so I also have alot of CPU years that are not
reflected in the Primenet stats.
I don't mean to say that you have no chance of finding a prime with a slower
machine. It just takes longer to check an exponent is all. It's that
length of time that I'm referring to, not your chances. I won't go into my
whole opinion on this again...for anyone interested, read the list archives
and you'll see my opinion on this spelled out quite well.
I *know* some people will disagree, so save your breath; realize that I
already know it's controversial and leave it at that.
> GIMPS is a cooperative effort, and frowning on people with shower
> machines is not friendly, and not in the spirit of GIMPS. We should be
> encouraging people to join the effort to find the next largest prime,
> and not discouraging them based on their processor type.
I appreciate every God blessed one of the < P-133's out there, even the
486's. I merely wish they'd focus on double-checking and factoring. Heck,
we're even reaching the point where double-checks on current exponents are
taking several months on P-100's and below...there *will* come a point at
which it no longer makes sense to use these machines even for double-checks.
I know, I know. Every machine in GIMPS is useful. But there is that fine
line where the usefulness is SOOOO minimal as to be nearly worthless. For
instance, ever wonder why there is no GIMPS client for an Apple ][, or for
an 8086 running DOS? Where's the Commodore client? I'm sure there are
millions of those lying in landfills all over. I'd bet 1 million
Commodore's would certainly be of great benefit to GIMPS. :-)
Of course, with the cost of electricity to run all those, you'd actually
SAVE money on power *and* run faster by getting a top of the line machine
anyway. And that's my point. At a certain point, the money you pay for
electricity just isn't worth it anymore. Donate those old machines to a
non-profit, take your generous tax write-off, get a new PC, save electricity
(new CPU's use less power anyway), run more iterations per second, etc.
Some people just hang on longer than others...I'm not criticizing that, I'm
merely asking that you think about doing factoring or double-checks for now.
Aaron
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
