| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Microbiome Immunity Project
Thread: Still only 1 % |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 70
|
|
| Author |
|
|
KerSamson
Master Cruncher Switzerland Joined: Jan 29, 2007 Post Count: 1684 Status: Offline Project Badges: 
|
Especially because MIP1 will be a long project, science optimisation will be more than welcome on both levels:
----------------------------------------1/ Optimisation (better efficiency) of the current science 2/ Strongly considering to develop a GPU-based science implementation Additionally, a support for Raspbian would be appreciated as well. I am still frustrated to have to observe on my machines the troubles caused by MIP1 and the poor efficiency of this project. Please learn from OET1 project! Cheers, Yves |
||
|
|
KLiK
Master Cruncher Croatia Joined: Nov 13, 2006 Post Count: 3108 Status: Offline Project Badges: 
|
Especially because MIP1 will be a long project, science optimisation will be more than welcome on both levels: 1/ Optimisation (better efficiency) of the current science 2/ Strongly considering to develop a GPU-based science implementation Additionally, a support for Raspbian would be appreciated as well. I am still frustrated to have to observe on my machines the troubles caused by MIP1 and the poor efficiency of this project. Please learn from OET1 project! Cheers, Yves If somebody can access data on over this link: https://boincstats.com/en/stats/-1/host/breakdown/os/ & post some answers. So many people can see that there's not much Raspbian OS present in BOINC computing World. It should be wise to switch Raspbian over to Android for Raspberry device & crunch on WCG.  |
||
|
|
[VENETO] boboviz
Senior Cruncher Joined: Aug 17, 2008 Post Count: 184 Status: Offline Project Badges: 
|
2/ Strongly considering to develop a GPU-based science implementation If i'm not wrong, MIP is based on rosetta code. So, no gpu app |
||
|
|
KerSamson
Master Cruncher Switzerland Joined: Jan 29, 2007 Post Count: 1684 Status: Offline Project Badges:
|
If a project should run for a decade, I would consider that some investment for a good and efficient implementation of the science would be more than meaningful.
----------------------------------------Cheers, Yves |
||
|
|
PabloMalaga
Advanced Cruncher Joined: Jul 7, 2011 Post Count: 51 Status: Offline Project Badges: 
|
Android???
|
||
|
|
Sgt.Joe
Ace Cruncher USA Joined: Jul 4, 2006 Post Count: 7844 Status: Offline Project Badges:
|
If a project should run for a decade, I would consider that some investment for a good and efficient implementation of the science would be more than meaningful. Cheers, Yves Android??? Not android, but I think he is alluding to a GPU implementation. Cheers
Sgt. Joe
*Minnesota Crunchers* |
||
|
|
PabloMalaga
Advanced Cruncher Joined: Jul 7, 2011 Post Count: 51 Status: Offline Project Badges:
|
I already know that it refers to GPU.
I asked if it is not interesting to accelerate MIP by implementing the WU for Android ?? Rosseta works for Android ... |
||
|
|
KerSamson
Master Cruncher Switzerland Joined: Jan 29, 2007 Post Count: 1684 Status: Offline Project Badges:
|
At this time, my primary concern regarding MIP1 is the science inefficiency.
----------------------------------------There are some troubles (in particular memory management) with MIP1 and it seems that nobody does really take care of it. For a 10 years project, I would expect that the code is efficient and clean. My bad feeling is that on the scientist side the poor performance is not seriously considered since the computational resource is free ! ... Cheers, Yves |
||
|
|
Sgt.Joe
Ace Cruncher USA Joined: Jul 4, 2006 Post Count: 7844 Status: Offline Project Badges:
|
I already know that it refers to GPU. I asked if it is not interesting to accelerate MIP by implementing the WU for Android ?? Rosseta works for Android ... Duly noted. Thanks for the clarification. At this time, my primary concern regarding MIP1 is the science inefficiency. There are some troubles (in particular memory management) with MIP1 and it seems that nobody does really take care of it. For a 10 years project, I would expect that the code is efficient and clean. For what it is worth, I am running MIP1 on a Q6600 with Vista Business with 2gb of memory. This particular machine configuration seems to run more efficiently than any of my faster machines running Linux or Windows 7. The only fly in the ointment is the need to exit BOINC before a shutdown to prevent the erroring of units. I don't have an explanation for it, but thought I would just mention it. Maybe it will provide a clue to making run better for everyone. Cheers
Sgt. Joe
*Minnesota Crunchers* |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
The only fly in the ointment is the need to exit BOINC before a shutdown to prevent the erroring of units. I don't have an explanation for it, but thought I would just mention it. Maybe it will provide a clue to making run better for everyone. Cheers I haven't had that issue...And I've been pretty negligent in that regard. I've also had the PC hibernating and sleeping with no issue at all. |
||
|
|
|