PrimeGrid
Please visit donation page to help the project cover running costs for this month
1) Message boards : Generalized Fermat Prime Search : High CPU usage (Message 156466)
Posted 25 days ago by Profile Crun-chiProject donor
Got it to work under Linux, thanks WezH for help.
Magic part of process is to copy libsleep.so to /usr/local/lib/. It is not working under home or even root directory.
Then you can start your program or service with
LD_PRELOAD="/usr/local/lib/libsleep.so" /path/to /your/app

I changed and compiled with value 60. If you dont make any changes in libsleep.c and compile it then you need to add this

'YIELD_SLEEP_TIME="yourvalue" LD_PRELOAD="/usr/local/lib/libsleep.so" /path/to /your/app

Thanks for solving problem. CPU usage and temperature lower significant and I dont see more then second slower result time.

Perfect!

2) Message boards : Aggie The Pew message board (Message 156434)
Posted 26 days ago by Profile Crun-chiProject donor
This time I will pass this challenge. Heat wave is at peak in my country... Best luck to all!
3) Message boards : Generalized Fermat Prime Search : High CPU usage (Message 156339)
Posted 34 days ago by Profile Crun-chiProject donor
Thanks for reply and solving solution, but compared to Windows this is still huge problem: under windows with GFN17 I have maybe 1-2% CPU utilization compared to 16-25% under Linux.
And gfsieve show same problem...
4) Message boards : Generalized Fermat Prime Search : High CPU usage (Message 156326)
Posted 35 days ago by Profile Crun-chiProject donor
Hi, I experiment with 1660Ti under Linux and noticed it eat one whole core. Under Windows that is no case.
So is any fix for this ( or maybe this is Linux/driver combination problem)

5) Message boards : Aggie The Pew message board (Message 156059)
Posted 55 days ago by Profile Crun-chiProject donor
And color is changed :)

6) Message boards : Aggie The Pew message board (Message 156048)
Posted 57 days ago by Profile Crun-chiProject donor
This is Anthony & me battling it out for five days. Then Crunchi blowing by both of us like a Space X rocket. 🚀



Ah , good old Ryzen(s) :) Just give him good cooler and there is no limit :)
7) Message boards : Aggie The Pew message board (Message 156043)
Posted 57 days ago by Profile Crun-chiProject donor
and soon will change color again :)

Nice. Is it a chameleon SR5 badge ? Ha


Yap, gold-amethyst :P
8) Message boards : Aggie The Pew message board (Message 156038)
Posted 57 days ago by Profile Crun-chiProject donor
and soon will change color again :)
9) Message boards : Aggie The Pew message board (Message 156025)
Posted 58 days ago by Profile Crun-chiProject donor
Anthony and I are neck-and-neck:
https://statsgr.id/custom_user_graph.php?challenge_id=2022_4&user0=90031&user1=386066

I have been following our toggle back and forth every few hours. LOL !

Last one to report a unit wins. That is where this is headed. :-)


https://statsgr.id/custom_user_graph.php?challenge_id=2022_4&user0=90031&user1=386066&user2=50683

I dont think I will catch you... but I try :)
10) Message boards : Number crunching : Strange timings on AMD 3700X (Message 156003)
Posted 60 days ago by Profile Crun-chiProject donor
It seems that when you have more than 4 workers, your L3 cache usage has exceeded your 32 MB of L3 cache.
So your 8 workers are causing the memory subsystem to thrash and all memory requests are going to slow memory, instead of fast L3 cache.

I have no problem with that, and I use 4 workers. But if as you say memory subsystem "fail" why then few cores are double faster then others?


Next 10 posts
[Return to PrimeGrid main page]
DNS Powered by DNSEXIT.COM
Copyright © 2005 - 2022 Rytis Slatkevičius (contact) and PrimeGrid community. Server load 0.61, 0.98, 0.85
Generated 19 Aug 2022 | 5:26:39 UTC