v2.2 Client -- Algorithmic Upgrade for All

**jjjjL** · 12-11-2004, 09:37 PM

News on the homepage.
http://www.seventeenorbust.com/

Download here:
http://www.seventeenorbust.com/download/

The speculation was correct. This version is a general release for all processors with the improvements from v2.0

Remember, install over the old client to keep your current test and username. Don't uninstall the client before upgrading.

Post your comments and benchmarks below. Thanks and enjoy.

Cheers,
Louie

**ceselb** · 12-11-2004, 10:23 PM

Great news. This will be interesting.

**IronBits** · 12-11-2004, 10:44 PM

Do the folks using the 'other' Intel P4 enhanced version need to upgrade?
How about the Dual Xeon folks? Any help, or keep running what we got?

For us AMD XP2600 types, I am getting an impressive 1,000,000+ cEMs/sec

As fast as one of my Xeon 3.6 logical processors.

Nice work

**Ken_g6[TA]** · 12-12-2004, 12:02 AM

Good news: On an old Celeron 667MHz (Intel just couldn't admit to having a devilishly fast (slow?) processor :P)

Before: 151033 cEM/s
After: up to 178000 cEM/s!

Better news: The new client successfully flushed to my Queue, which successfully flushed to your server.

I'm going to have to try replacing PRP with SB for my smaller prime finding. Louie (or Dave), I'd still love to get permission from you to release the code that lets me do this.

Thanks for the great new client!

**jamroga** · 12-12-2004, 02:39 AM

Client Timmings Old and New
Red Hat 9.0
Linux 2.4 Kernal

Pentium IV Celeron 2.4 Ghz

Old 2.0 384 sec
New 2.2 348 sec
Ratio ~ 1.1 x

Pentium III 600 Mhz

Old 1.2.5 1279 sec
New 2.2.0 831 sec
Ratio ~ 1.54 x

**em99010pepe** · 12-12-2004, 05:20 AM

AMD 1100 Mhz
Before: 350000 cEM/s
After: 570000 cEM/s

Carlos

**Matt** · 12-12-2004, 08:18 AM

Windows 2000 - Athlon XP2000+
Was: 400 k cEM/s
Now: 700 k cEM/s (+75%)

Other XP2000+ Box (worse/less RAM)
Windows 2000
Was: 300k cEM/s
Now (Peak): 750k cEM/s (+150%)
Now (Avg): 550k cEM/s (+83%)

FreeBSD boxes are all working, except for one which is doing a core dump (see other thread).

**Joh14vers6** · 12-12-2004, 10:22 AM

Louie, is also the last bug http://www.free-dc.org/forum/showthr...?threadid=7742 fixed in this version?

**Alien88** · 12-12-2004, 02:26 PM

Originally posted by Matt
Windows 2000 - Athlon XP2000+
Was: 400 k cEM/s
Now: 700 k cEM/s
Improvement: 75%.

Any idea when the FreeBSD client will be done? I can provide several boxes for testing or compiling on versions 4.9 4.10 and 5.2.1 if nessecary. Most of my crunchers are running FreeBSD so the sooner the better please!

It's available now.

**Alien88** · 12-12-2004, 02:27 PM

Originally posted by Joh14vers6
Louie, is also the last bug http://www.free-dc.org/forum/showthr...?threadid=7742 fixed in this version?

As far as I know it should be.

**jjjjL** · 12-12-2004, 06:28 PM

Originally posted by Joh14vers6
Louie, is also the last bug http://www.free-dc.org/forum/showthr...?threadid=7742 fixed in this version?

Yes.

**jjjjL** · 12-12-2004, 06:33 PM

Originally posted by Matt

Any idea when the FreeBSD client will be done? I can provide several boxes for testing or compiling on versions 4.9 4.10 and 5.2.1 if nessecary. Most of my crunchers are running FreeBSD so the sooner the better please!

FreeBSD client is online now. All clients are now online. Windows, Linux, FreeBSD, and even BeOS.

Cheers,
Louie

**IronBits** · 12-13-2004, 11:48 AM

SuperMicro X6DAE-G2, 1GB ram
Dual Xeon 3.6 GHz Nocona EM64T
Latest Windows XP64 2003 Server SP1
Running right at 5 Million cEMs/sec

On Windows 2000 it was 4 Million cEMS/sec

**vjs** · 12-13-2004, 03:06 PM

Is there an implementation of this new code for P-1's on athlonXP's?

**Vato** · 12-13-2004, 04:37 PM

Thanks for the hard work guys! Good Christmas present!

Runs sweet on 2K and XP on everything I've got from P3's, Athlons, P4's.

**IronBits** · 12-13-2004, 06:04 PM

Just and FYI... no scientific study, more of a 'first blush' ...
It appears the Windows XP is faster than Windows 2000 and SUSE 9.1 ...

Windows XP64 Server 2003 with Service Pack 1 just SCREAMS.

YMMV

Kudos and

to the coders for this release!

**Death** · 12-14-2004, 11:37 AM

using 2.2 version on 2k3 server sp1 can't connect to server. =((
server sb.pns.net port 1717
firewall turned off.
direct connection to internet, no proxies

or something.

dammit guys!!!!

C:\Program Files>ping sb.pns.net

Pinging sob.pns.net [69.9.160.186] with 32 bytes of data:

Reply from 69.9.160.186: bytes=32 time=264ms TTL=51
Reply from 69.9.160.186: bytes=32 time=244ms TTL=51
Reply from 69.9.160.186: bytes=32 time=467ms TTL=51
Reply from 69.9.160.186: bytes=32 time=588ms TTL=51

Ping statistics for 69.9.160.186:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 244ms, Maximum = 588ms, Average = 390ms

C:\WINDOWS>ping sbp.pns.net

Pinging sb-proxy.pns.net [69.9.160.188] with 32 bytes of data:

Reply from 69.9.160.188: bytes=32 time=217ms TTL=49
Reply from 69.9.160.188: bytes=32 time=197ms TTL=49
Reply from 69.9.160.188: bytes=32 time=194ms TTL=49

Ping statistics for 69.9.160.188:
Packets: Sent = 3, Received = 3, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 194ms, Maximum = 217ms, Average = 202ms

it all works with 69.9.160.188 ip but doesn't works with .186

what a suck!

**Ken_g6[TA]** · 12-14-2004, 03:05 PM

You might try my proxy, Death.

My replacement of PRP worked, too!

PII-400 running Linux testing 100000-digit numbers:
PRP time per test: 113 min
SB 2.2 time per test: 73 min!
Speedup: about 54.8%!

**Death** · 12-17-2004, 07:31 AM

Originally posted by Ken_g6[TA]
You might try my proxy, Death.

ofcourse I tried it some time ago, but think I cant use it now. just don't like java too much.

well, havin 2k3 server sp1 rc can give it another try =)) there's new version around.

but installer use sb.pns.net as server address. and a new user I suppose can't connect like me. I just did clean install and client can't connect to server

this is not working address!

C:\Program Files\SB>ping -a sb.pns.net

Pinging sob.pns.net [69.9.160.186] with 32 bytes of data:

Reply from 69.9.160.186: bytes=32 time=149ms TTL=49
Reply from 69.9.160.186: bytes=32 time=148ms TTL=49
Reply from 69.9.160.186: bytes=32 time=174ms TTL=49
Reply from 69.9.160.186: bytes=32 time=155ms TTL=49

Ping statistics for 69.9.160.186:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 148ms, Maximum = 174ms, Average = 156ms

this one is working

C:\Program Files\SB>ping -a sbp.pns.net
Ping request could not find host sbp.pns.net. Please check the name and try again.

C:\Program Files\SB>ping sbp.pns.net

Pinging sb-proxy.pns.net [69.9.160.188] with 32 bytes of data:

Reply from 69.9.160.188: bytes=32 time=289ms TTL=49
Reply from 69.9.160.188: bytes=32 time=356ms TTL=49
Reply from 69.9.160.188: bytes=32 time=322ms TTL=49
Reply from 69.9.160.188: bytes=32 time=321ms TTL=49

Ping statistics for 69.9.160.188:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 289ms, Maximum = 356ms, Average = 322ms

so there should be sb-proxy.pns.net address in installer

**Joh14vers6** · 12-20-2004, 07:07 AM

Originally posted by Joh14vers6
Louie, is also the last bug http://www.free-dc.org/forum/showthr...?threadid=7742 fixed in this version?

Originally posted by jjjjL
Yes.

I still see report that the bug is not solved. See http://www.free-dc.org/forum/showthr...threadid=8157.

**Matt** · 12-21-2004, 04:32 PM

Got around to updating my laptop,

AMD Athlon XP1500+ Mobile
256MB RAM
Windows 2000
Before: 200k cEM/s
After: 700k cEM/s (+250%)

I also found that by defragmenting my windows machines I've been able to get around a 200k cEM/s boost!!!

Thread: v2.2 Client -- Algorithmic Upgrade for All

Thread Tools

Rate This Thread

Display

v2.2 Client -- Algorithmic Upgrade for All

Client 2.2 Linux Some relative timmings

can't connect to server =(

Posting Permissions