Join PrimeGrid
Returning Participants
Community
Leader Boards
Results
Other
drummers-lowrise
|
1)
Message boards :
Number crunching :
2021 Challenge Schedule
(Message 146747)
Posted 891 days ago by Honza
Maybe PG could set up a drop on / drop off challenge, where ALL work done on SoB could be counted. Then those really favouring SoB and speaking about a year long challenge, can continue competing with those feeling and favouring SoB a lot, for the entire year, while those needing to do something different from time to time, can drop on / drop off as they suits. If PG incist that there needs to be a SoB challenge and it somehow interfears with the challenge schedule, to have an unofficial 2021 year scoreboard, counting the effort done on SoB over the year, then maybe an unofficial year long drop in / drop off challenge for TRP could be put in its place.
Tour de SoB, all year long, hopefully with a prime!
|
2)
Message boards :
Number crunching :
Surpassing the progress.
(Message 146642)
Posted 894 days ago by Honza
What is too large?
On what hardware?
During the years on PG, I witnessed projects to progress and tasks growing.
But I also witnessed apps speed optimalization (AVX, AVX-512), usage of GPUs, multi-threading etc.
HW and SW is also progressing.
*if* it comes to that point (for example, only 1 k of SoB remains and is ridiculously advanced), I can imagine pausing such a project if even best HW can't do thre job in a reasonable task running 24/7.
But I can't imagine there would be *no* participants willing to run 24/7 on best HW to finally finishing SoB and get the eternal fame!
|
3)
Message boards :
Number crunching :
Badges III
(Message 146623)
Posted 895 days ago by Honza
Another dirty diaper 
|
4)
Message boards :
Number crunching :
Running PSA
(Message 146446)
Posted 899 days ago by Honza
Please refer to Welcome to the Project Staging Area (PSA) for starter.
|
5)
Message boards :
Number crunching :
Why are GPU-capable subprojects also run on CPUs?
(Message 146207)
Posted 902 days ago by Honza
It is not always true that a CPU is slower than a GPU.
Sometimes, they are about the same performance.
Sometimes, GPU in a notebook is slower then CPU in the same computer.
Sometimes, one can even ask question: why are some GPUs running tasks that a CPU can handle better?
Note that apps are being developed and sometimes a new version gets considerable performance boost. See latest CPU Genefer for example.
Note that CPU can do 16 or more tasks at once using 150W, GPU usually does only one with 250W for example.
It is not only about speed, but overall performance or performance per watt.
For some participants, running tasks on CPUs is the only way to get their badge.
I guess there are plenty of reasons for running CPU or GPU or both.
It is an option, not mandatory.
EDIT: Mike's typing was faster :-)
|
6)
Message boards :
Number crunching :
Badges III
(Message 146110)
Posted 904 days ago by Honza
Wow, it was fast.
Congrats!
|
7)
Message boards :
Number crunching :
Finally, after 3 years of waiting...
(Message 145324)
Posted 920 days ago by Honza
Congrats!
|
8)
Message boards :
Generalized Fermat Prime Search :
Genefer 3.3.5 testing
(Message 145075)
Posted 930 days ago by Honza
Yves, it there a reason why Linux benchmark uses different values than Windows?
Are there different b limits even if CPUs are the same?
While we are testing new version, it might be convenient to have the same benchmark for all OS platforms...if that is a possible way.
|
9)
Message boards :
Number crunching :
Rtx 30 series
(Message 145049)
Posted 931 days ago by Honza
Can you please post benchmark numbers like I did for 3090?
Times are much better for comparation across the board, even with latest/beta GFN CPU app.
|
10)
Message boards :
Problems and Help :
Assign tasks to specific cores automatically
(Message 144655)
Posted 946 days ago by Honza
Note that all LLR task will be run under same app, ie. llr2_1.0.0_win64_200814.exe
ProcessLasso can assign equal number of cores / affinity for each instance.
If you run 4 tasks on 16-core 3950X like I do, it will assign 4 cores to each task.
|
Next 10 posts
|
