A new Linux App is available from our Beta Test page.
This is a "switching" App as you already know from the 4.38 App.
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.
The SSE App was built with compiler settings that should use the SSE unit for most arithmetics (-mfpmath=sse). There might be a difference in speed overall in one or the other direction, but mainly it should serve two purposes: first we try to avoid the FPU exceptions by avoid using the FPU, and second it should improve the prefetching of the Hough code (actually enable parts of it). This means that the App should run faster and further reduce the variance in run-times between workunits compared to previous versions.
In addition the App was built with BOINC API as of May 7, which means that it should work properly with latest development clients.
The app_info.xml has entries for
420
421
424
427
431
435
438
449
If your current App version is not listed here, you'll have to add it manually.
BM
BM
Copyright © 2024 Einstein@Home. All rights reserved.
GNU/Linux S5R3 App 4.49 available for Beta test
)
My AMD X2 5000 swiched fine, but my AMD XP 3000 wrecked all WUs.
From stderr.txt:
393, 394, 395, 396, 397, 398, 399, 400, 401, 402, 403, 404, 405, 406, 407, 408, Detected CPU type 1
SIGILL: illegal instruction
Stack trace (5 frames):
../../projects/einstein.phys.uwm.edu/einstein_S5R3_4.49_i686-pc-linux-gnu_1[0x81946a1]
[0xffffe420]
../../projects/einstein.phys.uwm.edu/einstein_S5R3_4.49_i686-pc-linux-gnu_1[0x806b368]
/lib/libc.so.6(__libc_start_main+0xe0)[0xb7d8dfe0]
../../projects/einstein.phys.uwm.edu/einstein_S5R3_4.49_i686-pc-linux-gnu_1(shmat+0x55)[0x804b5c1]
Exiting...
cu,
Michael
RE: My AMD X2 5000 swiched
)
Too bad. Are there SSE2 instructions in the executable?
Does the CPU_TYPE_0 work for the moment?
BM
BM
RE: RE: My AMD X2 5000
)
Don't know and I don't have the time right now to look at the dissassemby.
Well I use the 4.38 standard app again, which is working well(with SSE).
cu,
Michael
It indeed did contain SSE2
)
It indeed did contain SSE2 instructions. I built a new SSE App and updated the archive (and the md5sum on the page).
Thanks for the report.
BM
BM
RE: It indeed did contain
)
You're welcome. :)
I have noticed a significant speedup on my X2 5000. I store all finished and cached WUs in a RDBMS and calculate the runtimes for future WUs.
One WU, startet with app 4.35(516 skypoints), was successfully finished with 4.49.
Expected runtime would have been ~21100 sec, but only took 18726 sec!
This WU is away from the fastest WUs( alpha=0.472553 with alpha in [0,1] )
Future WUs will be even longer, so I'm really curios.
I have two questions:
1. Is it possible, that the first version(with some SSE2 instructions) might be faster than the new one?
2. Does the new beta app have any advantages for non SSE CPUs?
Next I'll try the new beta with my XP 3000 again.
cu,
Michael
[edit] typo
RE: 1. Is it possible, that
)
Yes, it is possible, but I simply don't know.
That reminds me that I wanted to post some "reference workunit" for direct speed comparison. Will pick this up soon.
Not in speed. I still want to test the new BOINC API version, though, that's in both Apps, in particular with the latest Core Clients.
BM
BM
The new beta runs fine on my
)
The new beta runs fine on my XP 3000.
cu,
Michael
Switched from 4.35 to 4.49 on
)
Switched from 4.35 to 4.49 on my AMD Opteron 1210 cpu running SuSE Linux 10.3 and BOINC 5.10.45. Looks definitely faster but graphics is not working. It used to work in 4.35.
Tullio
At the moment I have no Linux
)
At the moment I have no Linux machine to try this on, but I did notice a small typo on the beta test page:
I'm sure anyone that digs around will figure out that you meant to say "published 2008-05-14", but there's bound to be someone that posts and asks if you're using julian dates now. ;)
Good luck with it! I hope to have a linux box back up and going soon, so I'll catch the next beta. :)
RE: The new beta runs fine
)
It would be really interesting to know if the SSE2 capable version (first release) you have running on your X2 shows a bigger speedup than the "fixed" SSE version you now have running on your XP. In other words, is the speedup due mainly to the SSE2 instructions or is there a speedup for SSE only machines as well.
Don't lose that SSE2 version - it might be quite valuable :-).
Cheers,
Gary.