It seams to me that this would be a way to run BOINC on a dual boot system and use the same BOINC queue for booth windows and linux. So if one would install BOINC on the windows partition and use wine to run BOINC under linux then IMHO it would be possible
Then you're really interested in a subject, there is no way to avoid it. You have to read the Manual.
Well 1st result returned, with compute time of 59% of that for Linux client! Credit is still pending though.
For some reason it won't start the next WU, even though there's 7 cached locally.
I'm just investigating with a Seti@Home WU or 2 for comparison.
Reading this thread prompted me to install wine and make the comparison Wine/Linux. The results are from a system with an ASUS A8V motherboard : ADM 64 3500+ (2200Mhz) - no overcloking. On an otherwise idle system, with X11 running and a web browser opened.
As the results show, the i686-pc-linux-gnu version of boinc / einstein is a very bad performer - it's a pity to waste that much CPU performance.
Anyone listening, at EAH ?
-rg-
GNU/Linux native results : ----------------------------------
(kami)roger:/ligo/boinc
$ ./boinc*gnu -run_cpu_benchmarks
2005-03-20 23:55:12 [---] Starting BOINC client version 4.19 for i686-pc-linux-gnu
...
2005-03-20 23:55:12 [---] Running CPU benchmarks
2005-03-20 23:56:13 [---] Benchmark results:
2005-03-20 23:56:13 [---] Number of CPUs: 1
2005-03-20 23:56:13 [---] 1173 double precision MIPS (Whetstone) per CPU
2005-03-20 23:56:13 [---] 2688 integer MIPS (Dhrystone) per CPU
2005-03-20 23:56:13 [---] Finished CPU benchmarks
....
(kami)roger:/ligo/boinc
$
---------------------------------------------------------
Wine results : ------------------------------------------
$ wine boinc_cli.exe -run_cpu_benchmarks
To pause/resume tasks hit CTRL-C, to exit hit CTRL-BREAK
2005-03-20 23:47:20 [---] Starting BOINC client version 4.19 for windows_intelx86
...
2005-03-20 23:47:20 [---] Running CPU benchmarks
2005-03-20 23:48:22 [---] Benchmark results:
2005-03-20 23:48:22 [---] Number of CPUs: 1
2005-03-20 23:48:22 [---] 2071 double precision MIPS (Whetstone) per CPU
2005-03-20 23:48:22 [---] 5485 integer MIPS (Dhrystone) per CPU
2005-03-20 23:48:22 [---] Finished CPU benchmarks
...
(kami)roger:/ligo/Program Files/BOINC
$
----------------------------------------------------------
I seem to recall a similar issue with the pre-BOINC SETI code. Linux users of all species were getting better results using a WIN emulator than running the native Linux version.
They are probably listening, but at the same time, this isn't the easiest problem to solve. I am not even sure if boinc supports giving different clients differnet binaries based on the arch of the system, instead of just the OS. So we will have to wait and see
> Anyone listening, at EAH ?
such things just should not be writ so please destroy this if you wish to live 'tis better in ignorance to dwell than to go screaming into the abyss worse than hell
I'm running an optimized boinc client for the amd64 arch and I'm getting results comparable to the wine et etc setup... PC over here is amd64 3500 @2200 mhz
2005-03-18 17:56:27 [---] Running CPU benchmarks
2005-03-18 17:56:27 [---] Suspending computation and network activity - running
2005-03-18 17:57:28 [---] Benchmark results:
2005-03-18 17:57:28 [---] Number of CPUs: 1
2005-03-18 17:57:28 [---] 2253 double precision MIPS (Whetstone) per CPU
2005-03-18 17:57:28 [---] 5017 integer MIPS (Dhrystone) per CPU
I changed few my linux machines from running linux client to wine.
unfortunatelly i noticed that sometimes (once per few days) boinc or einsten core hangs up upon work completion. Core does nothing, boinc neither.
Killing wine-preloader and starting boinc again results in work upload.
Work is fine and is granted credits, however claimed crdits are very variable
(see here).
> I'm running an optimized boinc client for the amd64 arch and I'm getting
> results comparable to the wine et etc setup... PC over here is amd64 3500
> @2200 mhz
>
> 2005-03-18 17:56:27 [---] Running CPU benchmarks
> 2005-03-18 17:56:27 [---] Suspending computation and network activity -
> running
> 2005-03-18 17:57:28 [---] Benchmark results:
> 2005-03-18 17:57:28 [---] Number of CPUs: 1
> 2005-03-18 17:57:28 [---] 2253 double precision MIPS (Whetstone) per CPU
> 2005-03-18 17:57:28 [---] 5017 integer MIPS (Dhrystone) per CPU
>
>
>
Hi !
That looks good !
I had thought of installing a pure64 distro on my system, so I have now a good reason to try. But looking at your results, it seems that the einstein's client performance has not (greeatly) improved. Am I right ?
My guess would be the microsoft windows compiler. It is actually pretty good, better than gcc usually anyways.
> There's something rotten in the land of boinc Linux clients ..... Does
> somebody know what compiler was used for the windows boinc client?
such things just should not be writ so please destroy this if you wish to live 'tis better in ignorance to dwell than to go screaming into the abyss worse than hell
It seams to me that this
)
It seams to me that this would be a way to run BOINC on a dual boot system and use the same BOINC queue for booth windows and linux. So if one would install BOINC on the windows partition and use wine to run BOINC under linux then IMHO it would be possible
Then you're really interested in a subject, there is no way to avoid it. You have to read the Manual.
Well 1st result returned,
)
Well 1st result returned, with compute time of 59% of that for Linux client! Credit is still pending though.
For some reason it won't start the next WU, even though there's 7 cached locally.
I'm just investigating with a Seti@Home WU or 2 for comparison.
Hi ! Reading this thread
)
Hi !
Reading this thread prompted me to install wine and make the comparison Wine/Linux. The results are from a system with an ASUS A8V motherboard : ADM 64 3500+ (2200Mhz) - no overcloking. On an otherwise idle system, with X11 running and a web browser opened.
As the results show, the i686-pc-linux-gnu version of boinc / einstein is a very bad performer - it's a pity to waste that much CPU performance.
Anyone listening, at EAH ?
-rg-
GNU/Linux native results : ----------------------------------
(kami)roger:/ligo/boinc
$ ./boinc*gnu -run_cpu_benchmarks
2005-03-20 23:55:12 [---] Starting BOINC client version 4.19 for i686-pc-linux-gnu
...
2005-03-20 23:55:12 [---] Running CPU benchmarks
2005-03-20 23:56:13 [---] Benchmark results:
2005-03-20 23:56:13 [---] Number of CPUs: 1
2005-03-20 23:56:13 [---] 1173 double precision MIPS (Whetstone) per CPU
2005-03-20 23:56:13 [---] 2688 integer MIPS (Dhrystone) per CPU
2005-03-20 23:56:13 [---] Finished CPU benchmarks
....
(kami)roger:/ligo/boinc
$
---------------------------------------------------------
Wine results : ------------------------------------------
$ wine boinc_cli.exe -run_cpu_benchmarks
To pause/resume tasks hit CTRL-C, to exit hit CTRL-BREAK
2005-03-20 23:47:20 [---] Starting BOINC client version 4.19 for windows_intelx86
...
2005-03-20 23:47:20 [---] Running CPU benchmarks
2005-03-20 23:48:22 [---] Benchmark results:
2005-03-20 23:48:22 [---] Number of CPUs: 1
2005-03-20 23:48:22 [---] 2071 double precision MIPS (Whetstone) per CPU
2005-03-20 23:48:22 [---] 5485 integer MIPS (Dhrystone) per CPU
2005-03-20 23:48:22 [---] Finished CPU benchmarks
...
(kami)roger:/ligo/Program Files/BOINC
$
----------------------------------------------------------
I seem to recall a similar
)
I seem to recall a similar issue with the pre-BOINC SETI code. Linux users of all species were getting better results using a WIN emulator than running the native Linux version.
They are probably listening,
)
They are probably listening, but at the same time, this isn't the easiest problem to solve. I am not even sure if boinc supports giving different clients differnet binaries based on the arch of the system, instead of just the OS. So we will have to wait and see
> Anyone listening, at EAH ?
such things just should not be writ so please destroy this if you wish to live 'tis better in ignorance to dwell than to go screaming into the abyss worse than hell
I'm running an optimized
)
I'm running an optimized boinc client for the amd64 arch and I'm getting results comparable to the wine et etc setup... PC over here is amd64 3500 @2200 mhz
2005-03-18 17:56:27 [---] Running CPU benchmarks
2005-03-18 17:56:27 [---] Suspending computation and network activity - running
2005-03-18 17:57:28 [---] Benchmark results:
2005-03-18 17:57:28 [---] Number of CPUs: 1
2005-03-18 17:57:28 [---] 2253 double precision MIPS (Whetstone) per CPU
2005-03-18 17:57:28 [---] 5017 integer MIPS (Dhrystone) per CPU
I changed few my linux
)
I changed few my linux machines from running linux client to wine.
unfortunatelly i noticed that sometimes (once per few days) boinc or einsten core hangs up upon work completion. Core does nothing, boinc neither.
Killing wine-preloader and starting boinc again results in work upload.
Work is fine and is granted credits, however claimed crdits are very variable
(see here).
Anybody experienced anything similar ?
> I'm running an optimized
)
> I'm running an optimized boinc client for the amd64 arch and I'm getting
> results comparable to the wine et etc setup... PC over here is amd64 3500
> @2200 mhz
>
> 2005-03-18 17:56:27 [---] Running CPU benchmarks
> 2005-03-18 17:56:27 [---] Suspending computation and network activity -
> running
> 2005-03-18 17:57:28 [---] Benchmark results:
> 2005-03-18 17:57:28 [---] Number of CPUs: 1
> 2005-03-18 17:57:28 [---] 2253 double precision MIPS (Whetstone) per CPU
> 2005-03-18 17:57:28 [---] 5017 integer MIPS (Dhrystone) per CPU
>
>
>
Hi !
That looks good !
I had thought of installing a pure64 distro on my system, so I have now a good reason to try. But looking at your results, it seems that the einstein's client performance has not (greeatly) improved. Am I right ?
Rgds.
-rg-
There's something rotten in
)
There's something rotten in the land of boinc Linux clients ..... Does somebody know what compiler was used for the windows boinc client?
Regards
Mark
My guess would be the
)
My guess would be the microsoft windows compiler. It is actually pretty good, better than gcc usually anyways.
> There's something rotten in the land of boinc Linux clients ..... Does
> somebody know what compiler was used for the windows boinc client?
such things just should not be writ so please destroy this if you wish to live 'tis better in ignorance to dwell than to go screaming into the abyss worse than hell