Time reportet is almost everywhere zero, but it took always about 13000s to complete. But luckily the workunits get validated though.
Is there anything i can do to fix the time reported?
Yeah, I posted about his too farther down. All my machines just went into a forced CPU benchmark, so I am going to see if that fixes it for some reason.
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
If you're going to pin this thread, wouldn't it make more sense to instead have a single post thread listing where to download the two or three versions aksof(he's mentioned an I36 executable for P4's but not provided a link) has provided and which platforms they support, and then only allow moderators to post. As choked up as this thread is, actaully finding the needed information is easier said than done.
If you're going to pin this thread, wouldn't it make more sense to instead have a single post thread listing where to download the two or three versions aksof(he's mentioned an I36 executable for P4's but not provided a link) has provided and which platforms they support, and then only allow moderators to post. As choked up as this thread is, actaully finding the needed information is easier said than done.
Fair enough, it has been quite a whirlwind.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
I examined the linux code closely. It's horrible.
The 'blind flying' type optimization is hopeless. I'm sorry...
Probably a better compiler would be able to increase its performance significantly.
Hello, Akosf. Your application is gret, my crunching times are half. Would you be able to send me, or post a link to linux binary code for albert application? Or your binary for windows ? Maybe we would be able to make something with Linux and OSX version. Thanx, gabberattack
Just like to say congrats to Akosf for his great work. Do we know what the 'official' view of these new apps is at the moment? (sorry if this is posted elsewhere but I couldn't find it)
akosf said he looked at the disassembly of the linux binaries, but they were different enough he wasn't able to figure out what was going on.
Bruce's promised new, official applications based on akosf's algorithm improvements, but they haven't done so yet. Mainly I suspect the delay is in translating the assembly back into C to allow crossplatform development.
akosf said he looked at the disassembly of the linux binaries, but they were different enough he wasn't able to figure out what was going on.
Bruce's promised new, official applications based on akosf's algorithm improvements, but they haven't done so yet. Mainly I suspect the delay is in translating the assembly back into C to allow crossplatform development.
Here's an update.
We've now incorporated Akosf's improvements into our source code. But we haven't started distributing this faster application yet, for a simple reason. We are worried that our project server might break with the increased upload/validation disk load since the work will be getting done faster when we begin distributing new apps to all users. So we're upgrading the disk controllers and should be ready for this increased load soon.
We are worried that our project server might break with the increased upload/validation disk load since the work will be getting done faster when we begin distributing new apps to all users.
I can certainly understand why! The S38 optimized Windows SSE application does a workunit in about 0.30 the time of the current science application. That is like getting all at once over 3 times as many crunchers! At least that is a happy problem for you! :)
Definitely, hehe ;) what a problem! Would be interesting which app is faster, akosf's S-38 or the officially optimized one. Maybe the new official app will get another speed-up since Akos's further development may be not implemented yet.
Live long and crunch ;-)
RE: I've tried to run
)
Yeah, I posted about his too farther down. All my machines just went into a forced CPU benchmark, so I am going to see if that fixes it for some reason.
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
If you're going to pin this
)
If you're going to pin this thread, wouldn't it make more sense to instead have a single post thread listing where to download the two or three versions aksof(he's mentioned an I36 executable for P4's but not provided a link) has provided and which platforms they support, and then only allow moderators to post. As choked up as this thread is, actaully finding the needed information is easier said than done.
RE: If you're going to pin
)
Fair enough, it has been quite a whirlwind.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
On my Athlon XP2800+ it's
)
On my Athlon XP2800+ it's running at around 0.96% per minute!!!!!!!!
RE: I examined the linux
)
Hello, Akosf. Your application is gret, my crunching times are half. Would you be able to send me, or post a link to linux binary code for albert application? Or your binary for windows ? Maybe we would be able to make something with Linux and OSX version. Thanx, gabberattack
Just like to say congrats to
)
Just like to say congrats to Akosf for his great work. Do we know what the 'official' view of these new apps is at the moment? (sorry if this is posted elsewhere but I couldn't find it)
Cheers
Stwato
akosf said he looked at the
)
akosf said he looked at the disassembly of the linux binaries, but they were different enough he wasn't able to figure out what was going on.
Bruce's promised new, official applications based on akosf's algorithm improvements, but they haven't done so yet. Mainly I suspect the delay is in translating the assembly back into C to allow crossplatform development.
RE: akosf said he looked at
)
Here's an update.
We've now incorporated Akosf's improvements into our source code. But we haven't started distributing this faster application yet, for a simple reason. We are worried that our project server might break with the increased upload/validation disk load since the work will be getting done faster when we begin distributing new apps to all users. So we're upgrading the disk controllers and should be ready for this increased load soon.
Director, Einstein@Home
RE: We are worried that our
)
I can certainly understand why! The S38 optimized Windows SSE application does a workunit in about 0.30 the time of the current science application. That is like getting all at once over 3 times as many crunchers! At least that is a happy problem for you! :)
RE: At least that is a
)
Definitely, hehe ;) what a problem! Would be interesting which app is faster, akosf's S-38 or the officially optimized one. Maybe the new official app will get another speed-up since Akos's further development may be not implemented yet.
Live long and crunch ;-)
Greetings, Santas little helper