[QUOTE=Greenbank]A couple of RLX 300ex blade servers filled with 24 x 800MHz PIII (1GB RAM, 2x40GB HDD) blades. Each for ~ 1500 UKP.

There are some others on ebay (.co.uk) at the moment.[/QUOTE]

What is required to support it? Card cage/backplane/power supply/software?

It is a very nice chunk of hardware for the price!

[QUOTE=Greenbank]A couple of RLX 300ex blade servers filled with 24 x 800MHz PIII (1GB RAM, 2x40GB HDD) blades. Each for ~ 1500 UKP.

There are some others on ebay (.co.uk) at the moment.[/QUOTE]
Thanks very much for this info - the nearest I had got to guessing was something related to Google mini-servers, so...
J

[QUOTE=R.D. Silverman]What is required to support it? Card cage/backplane/power supply/software?

It is a very nice chunk of hardware for the price![/QUOTE]

The 300ex is the chassis. That provides the power supply, card cage, etc. It has a couple of 24 port switches built in (each blade has 2 NICs, one normal and one management network).

Then each blade has a PIII 800MHz Intel CPU, 1GB RAM, 2 x 40GB HDDs.

You then install Linux or FreeBSD on each blade (easy to do via PXE booting or by imaging each HDD) and then off you go. I'll go the Linux route probably.

At first I'll be using emclient with a few modifications. I'll eventually finish my own daemon code that requests a chunk of work (could be ECM, msieve or proth/riesel sieving chunks) and processes away then reports back to the server.

Update 2006-06-14
 

I just uploaded the latest update. Only two new factors this time, one by Bruce Hebden and one by me.

The latter, 3,2,388+, was the hardest of the composites in Bob's original tables. It really wasn't that hard. It took perhaps 25 GHz-days to sieve and another couple of days elapsed (on 1.3GHz and 2.2GHz AMD boxes) for the post processing. If anyone wants assistance finding SNFS parameters for other numbers, please let me or Bob know. I've already posted parameters for another in the tables but I believe that one will be attempted by Alex Greenbank.


Paul

6,5,217- C128 splits like so...
P40=1823735034640986315557065898724759393343
P89=12011381721766604323829242993786150915778576324120361533873037260431024008016285902578779

using ECM

[QUOTE=bhebden]6,5,217- C128 splits like so...
P40=1823735034640986315557065898724759393343
P89=12011381721766604323829242993786150915778576324120361533873037260431024008016285902578779

using ECM[/QUOTE]

And I was about 80% done with NFS...........

[QUOTE=R.D. Silverman]And I was about 80% done with NFS...........[/QUOTE]

I will now do 6,5,208+ instead. Reserve this for me.

[QUOTE=R.D. Silverman]I will now do 6,5,208+ instead. Reserve this for me.[/QUOTE]
Are you getting close to finishing it yet?

Pascal Ochem has sent me four new factorizations, all by GNFS. They are:
[code]4-3,299 C114 2445714137082801373090998119302291310171. P75
4-3,327 C115 4620199770191425227895590065467983960445897. P72
6-5,227 C116 51767429024691544495768395279144194185028827038570847. P64
6+5,197 C119 1723073241361648743035560632038543528462152438294257869. P65[/code]
and Bruce Hebden found this by ECM:
[code]6-5,217 C128 1823735034640986315557065898724759393343. P89[/code]

There are now 59 remaining composites, some of which are really quite easy by SNFS.

The updated tables will be on my web site at [url]http://www.leyland.vispa.com/numth/factorization/anbn/main.htm[/url] Real Soon Now.

Paul

[QUOTE=xilman]Are you getting close to finishing it yet?

Pascal Ochem has sent me four new factorizations, all by GNFS. They are:
[code]4-3,299 C114 2445714137082801373090998119302291310171. P75
4-3,327 C115 4620199770191425227895590065467983960445897. P72
6-5,227 C116 51767429024691544495768395279144194185028827038570847. P64
6+5,197 C119 1723073241361648743035560632038543528462152438294257869. P65[/code]
and Bruce Hebden found this by ECM:
[code]6-5,217 C128 1823735034640986315557065898724759393343. P89[/code]

There are now 59 remaining composites, some of which are really quite easy by SNFS.

The updated tables will be on my web site at [url]http://www.leyland.vispa.com/numth/factorization/anbn/main.htm[/url] Real Soon Now.

Paul[/QUOTE]


I had forgotten to post:

6,5,208+ =
208 (16) 2081.39521.1583572827326468983681.
5985084698480115491844811522820611873023501918020033.
309282182609411608949867128462131525483255114323883005643300752347457

[QUOTE=R.D. Silverman;84703]
[/QUOTE]

Here is 3,2,347-

347 (1) 2083.
4855947384634671768060977183621.
4403759916866552748196847493889639085538765415650172351938630263.
81712502485882300013423372771859484748484530121263986594491882236851

I am working on 6,5,199-.

[QUOTE=R.D. Silverman;85753]Here is 3,2,347-

347 (1) 2083.
4855947384634671768060977183621.
4403759916866552748196847493889639085538765415650172351938630263.
81712502485882300013423372771859484748484530121263986594491882236851

I am working on 6,5,199-.[/QUOTE]

Is anyone but P. Ochem and me working on these?

I predicted that people would pick off the low hanging fruit then give up.......