[QUOTE=harlee;119382]Can you expain the memory usage for ECM testing? Is there a difference if I say I have 128MB of memory vs 768MB, are the bounds the same, speed difference, etc).[/QUOTE]

The ECMers here might be able to better answer your question. The bounds do not change for memory settings. More memory may run faster depending on the exponent and B1/B2 values. I don't know if 128MB is enough that you've achieved most of the speed gains available for the exponents and bounds currently being assigned.

Sorry - not very helpful.

Thats ok. I've allocated 512MB of memory and currently the "ECM on small Mersenne number" that I'm currently work on - M264643 - is only taking up 330MB on stage 2.

[QUOTE=harlee;119382]Can you expain the memory usage for ECM testing? Is there a difference if I say I have 128MB of memory vs 768MB, are the bounds the same, speed difference, etc). tks[/QUOTE]I have no idea about speed differences, but for the exponents currently being assigned (e.g. M1272233) when I let Prime95 take a generous amount of RAM (e.g. 1.5GB) I've not seen it take more than 795MB for ECM stage2. Some exponents take 795MB, some take 455MB, the occasional one will take less, but apparently 795MB is as much as Prime95 thinks is useful for this range.

[b]edit[/b]: wait... no... I lied.
With 1.5GB available, Prime95 takes 795MB.
With 2.0GB available, Prime95 takes 1340MB.
With 3.0GB available, Prime95 takes 1340MB.

So, the question is with 1536MB allocated to Prime95 (and presumably some of that taken up by a LL test in thread#2), why doesn't Prime95 take 1340MB for the ECM stage2?

Request for the next version -- when starting stage2 could Prime95 print a message to the effect of "max efficiency needs <1234>MB of RAM; <768>MB available; using <620>MB"

Have question about the My Account -> Summary page. How often is the Performance Stats update? Wondering as those stats don't match that stats given under the Workload section. Makes sense if the Performance Stats is based on the hourly update and the Workload is update in real time. Sorry for the question as I'm just trying to understand the page.

[QUOTE=Prime95;119341]That could well be correct. A 55-bit factor is worth half of a 56-bit factor which is worth half of a 57-bit factor. With TF-LMH you are finding many 28-48 bit factors which from the above formula you can see are pretty much worthless. If you assume the factors found are uniformly distributed over the 28 to 57 bit range (they're not, they're actually skewed to the smaller factors) and only the last 3 or 4 bit levels will have any significant contribution to your CPU total then you can see the the factors-found CPU total being a 100 times smaller than the factors-not-found CPU total is probably right.[/QUOTE]Mini-Geek's 3305 exponents give a GHz-days ratio between factor:nofactor of 1:140. If my calculations are right, across the 304,062 exponents I've TF'd, I get an overall expected ratio of 1:112

[QUOTE=Prime95;119341]If you assume the factors found are uniformly distributed over the 28 to 57 bit range (they're not, they're actually skewed to the smaller factors)...[/QUOTE]A distribution like this? :smile:
(attaching here cause I can't attach 2 things to the previous message)

Every once in a while I get this

[CODE][Nov 29 23:05] == Info: Closing connection #0
[Nov 29 23:05] RESPONSE:
[Nov 29 23:05] pnErrorResult=0
[Nov 29 23:05] pnErrorDetail=SUCCESS
[Nov 29 23:05] ==END==
[Nov 29 23:05]
[Nov 29 23:05] Sending result to server:
[Nov 29 23:05] URL: http://v5.mersenne.org/v5server/?v=0.95&px=GIMPS&t=ar&g=9d0bcf97d87daa9f2f1701b3b9bb189f&k=DE9F75561C0B71FDD26AFC7788D45806&r=0&d=1&ss=1981
[Nov 29 23:05] == Info: About to connect() to v5.mersenne.org port 80
[Nov 29 23:05] == Info: Trying 64.66.6.254...
[Nov 29 23:05] == Info: connected
[Nov 29 23:05] == Info: Connected to v5.mersenne.org (64.66.6.254) port 80
[Nov 29 23:05] => Send header: GET /v5server/?v=0.95&px=GIMPS&t=ar&g=9d0bcf97d87daa9f2f1701b3b9bb189f&k=DE9F75561C0B71FDD26AFC7788D45806&r=0&d=1&ss=19815&sh=21
[Nov 29 23:05] Host: v5.mersenne.org
[Nov 29 23:05] Accept: */*
[Nov 29 23:05]
[Nov 29 23:05] <= Recv header: HTTP/1.1 200 OK
[Nov 29 23:05] <= Recv header: Server: Microsoft-IIS/5.0
[Nov 29 23:05] <= Recv header: Date: Thu, 29 Nov 2007 23:05:47 GMT
[Nov 29 23:05] <= Recv header: X-Powered-By: ASP.NET
[Nov 29 23:05] <= Recv header: MicrosoftOfficeWebServer: 5.0_Pub
[Nov 29 23:05] <= Recv header: Connection: close
[Nov 29 23:05] <= Recv header: X-Powered-By: PHP/5.2.5
[Nov 29 23:05] <= Recv header: Content-type: text/html
[Nov 29 23:05] <= Recv data: pnErrorResult=11
[Nov 29 23:05] pnErrorDetail=Update t_users failed, user_id: 40
[Nov 29 23:05] ==END==
[Nov 29 23:05]
[Nov 29 23:05] == Info: Closing connection #0
[Nov 29 23:05] RESPONSE:
[Nov 29 23:05] pnErrorResult=11
[Nov 29 23:05] pnErrorDetail=Update t_users failed, user_id: 40
[Nov 29 23:05] ==END==
[Nov 29 23:05]
[Nov 29 23:05] PrimeNet error 11: Server database corrupt
[Nov 29 23:05] Update t_users failed, user_id: 40
[Nov 29 23:05] Visit http://mersenneforum.org for help.
[Nov 29 23:05] Will try contacting server again in 1 minute.
[/CODE]

It waits a minute and then corrects and runs fine. Strange.... but not overly worrying I hope?

As a nice little milestone.... if I'm not mistaken, everyone in this beta has managed to TF-LMH a 100M digit range... go team! :grin:

I set one of my instances to P-1 factoring and the other to TF from TF-LMH and they both seem to be TF'ing between the 2^63-65 range?

[QUOTE=Bundu;119557]I set one of my instances to P-1 factoring and the other to TF from TF-LMH and they both seem to be TF'ing between the 2^63-65 range?[/QUOTE]

I'm getting P-1 assignments ok. Do the server web pages show the computer as having the correct work preferences? Can you send me the prime.log file?

Anything wrong with the Server? Just got home and I can't get any work and I can't log into my account (Squeeky_Squirrel).