Message boards :
Number crunching :
What is current application
Message board moderation
Author | Message |
---|---|
Send message Joined: 19 Aug 09 Posts: 23 Credit: 631,303 RAC: 0 |
Please forgive a simple question from a set-and-forget user. I see a lot of talk here about 0.21 applications, and nbody applications. I appear to be running MilkyWay 0.19 only. Is there any benefit to getting later versions, and if so, how is it done? |
Send message Joined: 17 Feb 08 Posts: 363 Credit: 258,227,990 RAC: 0 |
Please forgive a simple question from a set-and-forget user. No need to change anything. In fact the "new" apps are worse and like 5 times slower than what they should be. The slowdown is basically because the usage of a crappy compiler and developing the new app on a OS that only 5% of all attached machines are running, instead of developing it on windows and using a proper compiler... I think in english, such practice is called "shooting yourself in the foot"... Join Support science! Joinc Team BOINC United now! |
Send message Joined: 8 May 10 Posts: 576 Credit: 15,979,383 RAC: 0 |
The slowdown is basically because the usage of a crappy compiler and developing the new app on a OS that only 5% of all attached machines are running, instead of developing it on windows and using a proper compiler... This isn't even remotely true. That was not the problem. Windows is an insufferable development platform, and MSVC is an absolutely terrible compiler. The entire program is just math. There's nothing OS specific that could possibly cause such a slowdown. It ended up being a lack-of-SSE2 problem, with many improvements which assumed SSE2. The slow also affected 32 bit Linux. The main problem was the lack of SSE2 in the math libraries, so x86_64 Linux was fine, and then non-SSE2 builds were the problem. |
©2024 Astroinformatics Group