GNU/Linux S5R3 App 4.38 available for Beta test
29 Feb 2008 14:43:03 UTC
Topic 13655
(moderation:
A new Linux App is available from our Beta Test page.
Actually it's not a new App at all. The package includes renamed versions of the Apps formaly known as 4.31 and 4.35 and a little wrapper program that switches between them based on the CPU fetures it detects.
If something goes wrong with the automatic switching (i.e. you get "signal 4" "illegal instruction" errors), placing a file named "CPU_TYPE_0" in the BOINC directory should run the generic (non-SSE) App even if the wrapper would detect SSE.
This is a beta test for the wrapper more than anything else. If successful it should enable us to deliver optimized Apps to the larger public that doesn't run "power"- or "beta"-Apps.
The app_info.xml has entries for
420
421
424
427
431
435
438
If your current App version is not listed here, you'll have to add it manually.
BM
BM
GNU/Linux S5R3 App 4.38 available for Beta test
)
Oh yeah - the switcher should be linked statically. Will fix this.
BM
BM
RE: Oh yeah - the switcher
)
Done.
Have another go!
BM
BM
RE: Since we're talking
)
There's still some room for improvements of the SSE code, I'll try that first. My rough guess is that SSE2 would gain less than 10% over the best possible SSE App.
However with the new way of feature-based App switching the "costs" (for the project) are lowered too, compared to the cumbersome mixed-linked Apps we had before.
BM
BM
RE: There will definitely
)
In the new code we tried to avoid double precision as much as possible, so we already can perform most calculations in SSE. In the two functions that take the most time there is not much left to the compiler to optimize. Benefits from double-precision vectorization, more registers etc. are actually pretty minimal (e.g. the current kernel loop only uses 5 of 8 xmm registers, there is simply no benefit from having twice as many or even more).
There are a few specific features of SSE2 that are helpful, but only if the instructions are carefully placed into the code, probably in assembler (inline assembler in the code or using some well-coded math library). The full-blown 64Bit/SSE2 experiment where I left most to the compiler was nothing less than disappointing.
BM
BM
RE: Hi, Your optimised
)
I'll do this in the next few hours.
BM
BM
You may want to try the new
)
You may want to try the new 4.49 Beta App.
BM
BM