Enigma@Home: reliable project

Message boards : Number crunching : Enigma@Home: reliable project

To post messages, you must log in.

AuthorMessage
Extra Ball

Send message
Joined: 5 Jul 09
Posts: 1
Credit: 503,915
RAC: 0
Message 1263 - Posted: 24 Sep 2009, 13:47:50 UTC

The current AF action (Autumn target) will finish soon
The Enigma project has proven it could handle a considerable increase of charge with no prob at all, during the period of the challenge (2 weeks)

As a single AF member with no reponsabilities, I would like to congratulate the Enigma Team members for their efforts (1st) plus the way this action was made possible (resources never failed)

We expected a server crash. We failed.
We will be back :)

GG guys




ID: 1263 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile TJM
Project administrator
Project developer
Project scientist
Avatar

Send message
Joined: 25 Aug 07
Posts: 843
Credit: 267,994,998
RAC: 0
Message 1264 - Posted: 24 Sep 2009, 23:26:50 UTC - in response to Message 1263.  

During the last couple of weeks the project server went through a real stress-test :-D

At some point it started having problems handling the workload during peak hours, when the throughput was almost 5 times higher than the average. I thought it was the hardware limit - the machine has never seen a load like this before (and the server is just an average desktop PC), but after couple of minor changes in the db layout everything went back to normal.

Awgly100 search space was pushed closer to the end. I guess it's time to start looking for other ciphertext(s) and to ask Stefan Krah what he plans to do next. http://www.cryptocellar.org has a large number of unbroken M3 messages - I already thought about adding one of them as a 'backup plan' in case of communication failure between BOINC and M4 servers.

I'm also thinking about 'plan B' for the current message - if everything else fails, perhaps the message could be broken by a brute force combined with scoring functions to automatically find best results. I think it can't be done on CPUs (the number of all combinations is way too large), but the computing power of the modern GPUs might be high enough to try.
M4 Project homepage
M4 Project wiki
ID: 1264 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile rgathright

Send message
Joined: 8 Jul 09
Posts: 2
Credit: 67,662
RAC: 0
Message 1265 - Posted: 25 Sep 2009, 14:20:57 UTC

GPU's eh? Did you know that ATI gave the SETI project a free video card and documentation after they submitted a request for help?
Overclock with the MSI G31M3-L and Intel E8600 3.33Ghz

Intel D865GLC Socket 478 Motherboard Review

Overclock your ASUS 1005HA netbook and crunch SETI
ID: 1265 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
fitz

Send message
Joined: 15 Apr 09
Posts: 31
Credit: 147,954
RAC: 0
Message 1268 - Posted: 28 Sep 2009, 8:21:54 UTC

I have an ATI card that i would be happy to try running stuff on...unfortunatly coding in open CL or CAL is a bit beyound me atm!
ID: 1268 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
quel

Send message
Joined: 19 May 09
Posts: 34
Credit: 32,923,471
RAC: 0
Message 1269 - Posted: 28 Sep 2009, 18:11:45 UTC - in response to Message 1264.  

It seems a lot of the M3 messages are just too short for much to work. However, from http://cryptocellar.web.cern.ch/cryptocellar/bgac/GArmy_messages.html there are certainly some longer messages.

Enigma message of 27.06.1941, Nr. 48, (WRMKX): 141 chars
Enigma message of 08.07.1941, Nr. 23, (KHLPT): 125 chars
Enigma message of 13.08.1941, Nr. 7, (KEJNQ): 158 chars

Assuming the messages aren't garbled in any way the longer the ciphertext the easier it should be to break right? That is to travel the easy road we should start with the longest ciphertext first.
ID: 1269 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile TJM
Project administrator
Project developer
Project scientist
Avatar

Send message
Joined: 25 Aug 07
Posts: 843
Credit: 267,994,998
RAC: 0
Message 1270 - Posted: 28 Sep 2009, 23:30:17 UTC - in response to Message 1269.  
Last modified: 28 Sep 2009, 23:56:36 UTC


Assuming the messages aren't garbled in any way the longer the ciphertext the easier it should be to break right?


Theoretically yes, but processing time needed for a single workunit (and the whole keyspace) increases very fast with the ciphertext lenght, the increase is closer to exponential than linear, so I think it's better to truncate the ciphertext to lenght which was proven to be breakable (~100 letters or even less).

It seems a lot of the M3 messages are just too short for much to work.


Most of messages longer than 40-50 letters are breakable with a reasonable number of restarts, at least these which I used for testing were. With some luck even a 20 letters text can be broken, but it requires app modification and very good dictionaries.


fitz wrote:
I have an ATI card that i would be happy to try running stuff on...unfortunatly coding in open CL or CAL is a bit beyound me atm!


Same here, I'm not a programmer - I can barely write c/c++ code, but I like to learn new things and GPU programming is really interesting. I'm going to try nVidia/CUDA first, because it seems easier for me, the cheapest used CUDA-compatible cards are almost free (8400/8500 series) and the available documentation is better (at least for now) than ATI's.
M4 Project homepage
M4 Project wiki
ID: 1270 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile plonk420

Send message
Joined: 5 Jun 09
Posts: 10
Credit: 8,627,630
RAC: 0
Message 1315 - Posted: 3 Nov 2009, 22:28:22 UTC - in response to Message 1264.  

Awgly100 search space was pushed closer to the end. I guess it's time to start looking for other ciphertext(s) and to ask Stefan Krah what he plans to do next. http://www.cryptocellar.org has a large number of unbroken M3 messages - I already thought about adding one of them as a 'backup plan' in case of communication failure between BOINC and M4 servers.


are these historical messages?
ID: 1315 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile TJM
Project administrator
Project developer
Project scientist
Avatar

Send message
Joined: 25 Aug 07
Posts: 843
Credit: 267,994,998
RAC: 0
Message 1317 - Posted: 7 Nov 2009, 12:52:51 UTC

Yep, there are a lot of unbroken german army messages from 1941.

M4 Project homepage
M4 Project wiki
ID: 1317 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Daniel

Send message
Joined: 9 Dec 07
Posts: 5
Credit: 6,437,398
RAC: 0
Message 1359 - Posted: 16 Nov 2009, 10:31:29 UTC

Kickstart anyone?
ID: 1359 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile [AF>France>Aquitaine>Cote-Adour-et-Gaves]Bernard 64250

Send message
Joined: 24 Sep 09
Posts: 18
Credit: 4,428,367
RAC: 0
Message 1367 - Posted: 16 Nov 2009, 23:13:24 UTC
Last modified: 16 Nov 2009, 23:14:12 UTC

Is the server down ? I still have a few WUs waiting for upload.
ID: 1367 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile TJM
Project administrator
Project developer
Project scientist
Avatar

Send message
Joined: 25 Aug 07
Posts: 843
Credit: 267,994,998
RAC: 0
Message 1369 - Posted: 16 Nov 2009, 23:28:08 UTC - in response to Message 1367.  

Yep, it`s down. Probably won`t be up for the next 12 hours due to hdd failure.
M4 Project homepage
M4 Project wiki
ID: 1369 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile mdoerner
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 30 Jul 08
Posts: 202
Credit: 6,998,388
RAC: 0
Message 1374 - Posted: 18 Nov 2009, 17:25:14 UTC - in response to Message 1369.  

Good, now I can save some $$$ on my electric bill until we're back up again....;-)

Mike D
ID: 1374 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote

Message boards : Number crunching : Enigma@Home: reliable project




Copyright © 2024 TJM