Recent spee preject speed increase

[nuutti]

When I look at the last 24 h top 10 I see 5 / 10 participant that
have joined during last week.
Are we gaining momentum and big boys are joining ?

And is the recent huge jump in Systemwide cExpMod rate real ?

If project total speed is real then we will reach 3,000,000 milestone very soon !

I think that it would be good for project that we can find one prime as soon as possible. New potential participants can see that we are doing progress. so I think that we should check as many k's as possible up to 3,000,000 before going higher in n.


My two cents,

Nuutti

[Alien88]

Originally posted by nuutti
When I look at the last 24 h top 10 I see 5 / 10 participant that
have joined during last week.
Are we gaining momentum and big boys are joining ?

And is the recent huge jump in Systemwide cExpMod rate real ?

If project total speed is real then we will reach 3,000,000 milestone very soon !

I think that it would be good for project that we can find one prime as soon as possible. New potential participants can see that we are doing progress. so I think that we should check as many k's as possible up to 3,000,000 before going higher in n.


My two cents,

Nuutti

With the addition of teams, we have picked up members from other places who had been waiting for team stats to be created. The speed increase is also due to the newer clients having a major performance increase compared to the older ones. The current plan is to test up to 3,000,000 and then we will move on to a new n value.

[MAD-ness]

The recent changes to the client (v .98 and then v.99, with Linux and FreeBSD being caught up now) made a huge difference to people who are active in distributed computing teams and projects. The hard-core kinds of guys. :)

I imagine that things will continue growing from here. There seems to be a certain point where a DC project either peters or out or the client gets good enough/friendly enough for it to continute growing and pick a sort of critical mass.

I would imagine that TPR, atleast, will be about double its current size (in terms of mhz) by xmas. Really rough guess. If other DC teams decide to compete seriously here, that could happen with 3-4 of them.

Either way, things should continue heating up. :)

[Firebirth]

Actually i think that it is kind'a ad, that we are ending 27653 at 3000000... I would like to see it progress, if we don't find a prime under 3G... couldn't it be an option in the config file to choose, which of the sequences you want to work on?

- I truly believe that there is a prime somewhere out there; I feel, that we would be wasting our work, if we completely just stop at a certain point - although I think it is a good thing to start working on something that may be grasped in reasonsable time!

Mevs Tarje

[nuutti]

Of course there is a prime, but It can be in range
n = 20,000,000 -> 40,000,00 or n=40,000,000 -> 80,000,000.

Yves Gallot has estimated that we have 50% chance to find
all remaining primes when we have checked all candidates up to
n = 10^13 = 10,000,000,000,000 so it is very likely that this project is up and running 2010. Lets say that last 3 -4 are very very hard to find.

You can find his paper from :
http://perso.wanadoo.fr/yves.gallot/papers/weight.html


Yours,

Nuutti

[Mystwalker]

I don't think the work was a waste - it's just one part of the whole work.
We also don't habe to start the other k's from scratch, as there were people before who did some maths. And maybe there will be another project that can continue with our results - assuming it won't be us who restart that k. ;)
Plus, it is more likely to find a prime with a lower n AFAIK. So it should be ok to keep the n's for the k's even (at least those 3 SB is assigned to), alas we would return to 27653 when both other k's hit the 3M.

But you're right, an option to choose the k you want to work on would be nice.