![]() |
|
|
#1 |
|
1DA116 Posts |
* Trial factoring *
With primetools mfloop.py it is possible to stop the loop: 1. first stop the script; 2. compute all assignments until the program terminates with the error "no assignments found"; 3. invoke the script with arguments ... -n 0 -t 0 to report the last results. * PRP * With gpuowl primenet.py the procedure is a bit different: we have to report the last result by hand via the web page. So I propose a little change to primenet.py : - Add a command line argument that say how many assignments we want to fetch (with possibility to set to 0 to only report results). Last fiddled with by SELROC on 2019-03-31 at 07:41 |
|
|
|
#2 |
|
"Mihai Preda"
Apr 2015
22×3×112 Posts |
In GPUowl primenet.py, added the optional --tasks which allows to specify the number of fetch-ahead tasks. If not specified, this is 2 for PRP and 12 for P-1.
|
|
|
|
|
|
#3 |
|
10000100000102 Posts |
|
|
![]() |
| Thread Tools | |
Similar Threads
|
||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| Automatic fetch of Trial Factoring work for GPU mfakt* | LaurV | GPU to 72 | 81 | 2020-12-02 05:17 |
| Work Fetch - External Command | 0PolarBearsHere | MISFIT | 31 | 2017-08-30 20:44 |
| x86 ASM loop control: JUMP vs LOOP | ewmayer | Programming | 23 | 2012-04-29 21:13 |
| Stopping New Work | dmaine | Information & Answers | 1 | 2011-07-05 22:30 |
| v25.7 Destroys PrimeNet v4 work-in-progress | Freightyard | Software | 9 | 2008-11-08 18:01 |