Attention RC5 members using K6 through K6-3.

[Jator]

I'm calling for everyone who uses a K6 through K6-3 chip to participate in Distributed.net's challenges to look at the return you are getting from your processor as oposed to what you could be getting if you participated with OGR.

With RC5, you are only getting about 60% of the output from the same clocked PentiumII/Celron or above processor. If you switched these processors to OGR, you would get 95% of the work that would be done by the same clocked Intel chip. That's a pretty big jump.

Also, think of this: Although moving the cip from RC5 to OGR will take some production away from the RC5 effort, you are making up for it and then some for the OGR effort.

One last thing to think about. With over 1/3 of the RC5 keyspace checked, each day bring us closer to the possibility of the project ending. Might as well get a head start on the next big project.

If you check the daily stats page for DIstributed.net, you will notice I am participating with both OGR and RC5, thus I feel that I am not playing favoritism here. Just trying to be practical. 😉

If you want to check for yourself, look at the speed page provided by Mika: Speed Page

Jay

 

[Bleep]

Jator:
I have been thinking of doing just that. Not because of the K6 processor although I have one machine with a k6 500 that would do pretty good OGR. Lots of things going on with the rc5 bunch not very appealing to me. stats Bleep@pwhipped.com
Bleep

 

[Jator]

Bleep,

Glad to hear you are thinking of it. Feel free to use my proy so that you can get real time stats (OGR and RC5). I dump to Mika, so no loss there.

Sorry to hear you are not happy with the RC5 team. Things have been pretty smooth in my eyes for almost a month now.

Jay

 

[Viztech]

Yes, there is no doubt that those K6s are best at OGR. I'll get 1.7 gig worth of K6s moved over to OGR this week from those pesky RC5Crackheads.

viz

 

[Deeko]

Correct me if I'm wrong, but we don't try to recruit you do we? 😛

 

[Bleep]

Deeko:
Was your question addressed to me?? And if so who is the WE that you are talking about. As far as team membership is concerned I have been cracking for Anandtech since about the 5th day it started quite some time ago. In fact the client I am using is very old. 2.7106.436
Bleep

 

[PeterN]

The OGR-project for TA seems to get up to speed.

It will take some power from the RC5-project, but it will benefit TA at least. 😛😛

Sorry to hear you're feeling uncomfortable with the RC-5 'bunch', Bleep.🙁

Do you want to talk about it by PM or mail?
Maybe something can be cleared?
I hate to see such a valuable TA-member being sad.

Peter.

 

[mindless]

while were on the topic of OGR, how does one stop the client from downloading so many work units? When I switched this p2 400 over from RC5 almost a week ago it downloaded 23 work units, it had completed 9 of them so far and I wanted some OGR stats so I ran dnetc -update, it flushed all 9 but downloaded 9 more, I cant wait a month for stats for OGR to get updated :|

heres my ini:

[parameters]
id=phf@stayawayfromme.com (no thats not my real email)

[misc]
project-priority=RC5=0,DES=0,OGR,CSC=0

[triggers]
restart-on-config-file-change=yes

[networking]
autofindkeyserver=no
keyserver=proxy80.teamanandtech.com

[rc5]
randomprefix=237

[logging]
log-file=rc5.log
log-file-type=fifo

 

[vss1980]

try modifying or adding this:

[ogr]
fetch-workunit-threshold=xxx

where 'xxx' is the number of WU's you want buffered.

Edit: I've heard rumours that changing your ID to anandtechrc5@yahoo.com makes it go faster 😉

 

[mindless]

Thanks, now I just have to wait for me to go through these 23 work units..

anandtechrc5@yahoo.com will make my client go faster? if you can prove it with stats I will change it 😉

 

[dennilfloss]

Go Team AnandTech! 🙂

 

[RaySun2Be]

Mindless,
Set the OGR threshold to contain several days worth of WUs, just in case of connection or keyserver problems. Then set the following to flush/fetch work as it is completed, that way you won't have to wait for the threshold to be met before updates:

[buffers]
frequent-threshold-checks=3


Ray