Is there a new faster application available?

I've found by accident that user NetMonitoring is delivering results with a runtime of 400 seconds each.
Task info shows he's running application version 0.71.
Official application version is 0.07.

Anyone any idea where to get this faster app?

Probably from that user. Would be a custom app.
____________
Doesn't expecting the unexpected make the unexpected the expected?
If it makes sense, DON'T do it.

If you were using Linux, petros or I could help you out... ;-)

____________
mic.

Has anyone made a faster app for XP?
____________
Doesn't expecting the unexpected make the unexpected the expected?
If it makes sense, DON'T do it.

Well, NetMonitoring was running the app on a Windows XP machine
according to the Computer entry of the task id.

Obviously there *is* a faster app for XP out somewhere ...

Well I suppose I should add: One that someone is willing to share?
____________
Doesn't expecting the unexpected make the unexpected the expected?
If it makes sense, DON'T do it.

Well I suppose I should add: One that someone is willing to share?

Thanks for the links.

How do I figure out which I have? I would guess sse or sse2. My pc's a P4 xp.
____________
Doesn't expecting the unexpected make the unexpected the expected?
If it makes sense, DON'T do it.

Thanks for the links.

How do I figure out which I have? I would guess sse or sse2. My pc's a P4 xp.

Thanks for the links.

How do I figure out which I have? I would guess sse or sse2. My pc's a P4 xp.

Thanks sandro, your links are much appreciated.

I've only just started to try them out and can see everything ticking away faster. What was indicating at over 1 hour to complete is now showing at under 40 minutes. Not very scientific I know, but they seem to be doing the trick...

____________

[quote]Thanks for the links.

Let me know when you find one of these fast/optimised MW clients for Windows and the link, please.

I think I'm missing something. Boinc wants to keep downloading the official app. I added an app_info file I converted from milksops app. And now I'm getting "no work", so I can't tell if I did it right.

Here's what I have in that file:
<app_info>
<app>
<name>astronomy</name>
</app>
<file_info>
<name>astronomy_sse2_app.exe</name>
<executable/>
</file_info>
<app_version>
<app_name>astronomy</app_name>
<version_num>1</version_num>
<file_ref>
<file_name>astronomy_sse2_app.exe</file_name>
<main_program/>
</file_ref>
</app_version>
<app_info>
____________
Doesn't expecting the unexpected make the unexpected the expected?
If it makes sense, DON'T do it.

banditwolf, there is an app_info file for all SSE, SSE2 and SSE3 in the downloads (one download has it missing, but all can be found in sandro's links - you don't have to make any app_info files)

I'm using all 3 and they all seem to be working OK.
____________

banditwolf, there is an app_info file for all SSE, SSE2 and SSE3 in the downloads (one download has it missing, but all can be found in sandro's links - you don't have to make any app_info files)

banditwolf, there is an app_info file for all SSE, SSE2 and SSE3 in the downloads (one download has it missing, but all can be found in sandro's links - you don't have to make any app_info files)

I went with the 2nd sse2, only the exe file. The first is a .rar file. I don't know how to extract that one.

http://www.file-upload.net/download-1319985/MW_SSEV2.zip.html
SSE2+SSE3
http://www.file-upload.net/download-1320653/MW.rar.html

these I get what looks to be an error saying too many requests.
____________
Doesn't expecting the unexpected make the unexpected the expected?
If it makes sense, DON'T do it.

Cripes! get winzip or winrar quick - WUs that took over an hour an hour ago are now rolling off at under 20 minutes on my core 2 duo.

http://www.file-upload.net/download-1319985/MW_SSEV2.zip.html
SSE2+SSE3
http://www.file-upload.net/download-1320653/MW.rar.html

these I get what looks to be an error saying too many requests.

Gipsel is working on a GPU-app für ATI cards. he is searching volunteers for testing his GPU-app under win64 with ATI cards. if some wants to test it, please pm gipsel in the p3d board

If the GPU-app is working a WU could be crunched in 20sek (but 1 cpu core would probably be used also

could anyone reupload the SSE2 and SSE3 versions? and ho it looks witth SSE 4.1? thx for answer

I still get 'no new work'. Could it be from the 8 that I aborted to get the app correct? Guess i have to wait till they are out of the system.
____________
Doesn't expecting the unexpected make the unexpected the expected?
If it makes sense, DON'T do it.

I still get 'no new work'. Could it be from the 8 that I aborted to get the app correct? Guess i have to wait till they are out of the system.

I finished a stripe_82 in 38 min down from 82-85 min. I'm using the sse2 - 2nd down version. The end of running ~ the last 5% doesn't tick of smoothly, it jumps .5-.6% every 20-30 seconds.

edit: Actually progress is a little jumpy most of the time. But is still accurate when it moves.
____________
Doesn't expecting the unexpected make the unexpected the expected?
If it makes sense, DON'T do it.

SSE2+SSE3
http://www.file-upload.net/download-1320653/MW.rar.html

Links are back to live...

I downloaded and installed the SSE3 app: it's working fast as hell.

Many thanks to P3D. You are doing a great job!

Senilix

The zip and rar files made available earlier by sandro (except for the one that had no app_info file) can be downloaded from one of my web sites here;

www.arizmoon.com

The apps are...... amazing!!! You'll be over the moon with them ;)

Thanks again for the original links sandro, and to Gipsel and twodee for making them.
____________

Thanks to Sandro for posting the links, and Gipsel and twodee for working on the apps, and Ice for adding an additional dl location. It's a great improvement.

I'll try the sse3 app later and see if it works on my system.
____________
Doesn't expecting the unexpected make the unexpected the expected?
If it makes sense, DON'T do it.

Thanks to Sandro for posting the links, and Gipsel and twodee for working on the apps, and Ice for adding an additional dl location. It's a great improvement.

Try it here, if you want

http://www.file-upload.net/download-1311757/astronomy_sse2_app.exe.html

This one doesn't seem to work on my machine

Hello, Please excuse my ignorance; I downloaded the app, now what do I do with it??

Thanks!!

Beau

First. Unzip it.

Copy the files unziped in your project folder milkyway.cs.edu_milkyway

Stop BOINC (rememer to stop the daemon in the tools bar with the right button and selecct 'stop') and then restart BOINC. Only that.
____________
Logan.

BOINC FAQ Service (Ahora, también disponible en Español/Now available in Spanish)

Of course it would be that simple, yet I had to ask!! I will give it a try and see what happens :-)

Thanks for the assistance!!

Beau

Of course it would be that simple, yet I had to ask!! I will give it a try and see what happens :-)

Thanks for the assistance!!

Beau

Of course it would be that simple, yet I had to ask!! I will give it a try and see what happens :-)

Thanks for the assistance!!

Beau

If you need any help, don't worry for ask me. (sorry for my bad english).

Are there any links to a Linux app? The links given seem to be all for Windows :(

Thanks in advance..

Are there any links to a Linux app? The links given seem to be all for Windows :(

Thanks in advance..

www.arizmoon.com

The apps are...... amazing!!! You'll be over the moon with them ;)

Can anyone tell me which DL to use for 64bit and which one to use for 32bit windoz?
____________
A clear conscience is usually the sign of a bad memory

Are there any links to a Linux app? The links given seem to be all for Windows :(

Thanks in advance..

Can anyone tell me which DL to use for 64bit and which one to use for 32bit windoz?

The zip and rar files made available earlier by sandro (except for the one that had no app_info file) can be downloaded from one of my web sites here;

www.arizmoon.com

The apps are...... amazing!!! You'll be over the moon with them ;)

Thanks again for the original links sandro, and to Gipsel and twodee for making them.

The zip and rar files made available earlier by sandro (except for the one that had no app_info file) can be downloaded from one of my web sites here;

www.arizmoon.com

The apps are...... amazing!!! You'll be over the moon with them ;)

Thanks again for the original links sandro, and to Gipsel and twodee for making them.

ICE I'm a bit confused on what file I should have. I have an AMD x2 6400 and it has SSE,SSE2,SSE3. I think I need to go with SSE2 so that the other system can run it without my getting confused about who has what or where. Oh and I don't use winrar... sigh

Well that was fast. Nothing like having a ton of computer errors and it sending them in so fast that it was hard to stop it. Hard to detact and thankfully put my backup I made back on that I had made just before I tried this little experiment.. EXTRA SIGH

Are results valid comparing to official app?
Also - aren't You affraid of next "no support for Milky..." as before?

Faster applications returns incorrect results.

Original app:
fitness: -3.046272445362277

"Faster app":
fitness: -3.046289346362010

This is not good. The faster application is throwing out garbage results. These results will not be counted and you will not get credit for them.
____________
Dave Przybylo
MilkyWay@home Developer
Department of Computer Science
Rensselaer Polytechnic Institute

Faster applications returns incorrect results.

Original app:
fitness: -3.046272445362277

"Faster app":
fitness: -3.046289346362010

This is not good. The faster application is throwing out garbage results. These results will not be counted and you will not get credit for them.

I have just checked the results which are viewable on my rigs and all seems to be OK here. I hope I can continue like that.

I can certainly see why Dave is concerned as this puts unnecessary strain on the servers without valid results coming back, meaning these WUs need to be reissued.

Which of the faster apps gives out garbage results?
SSE 2 or SSE3? Or all? *scratches head*

Faster applications returns incorrect results.

Original app:
fitness: -3.046272445362277

"Faster app":
fitness: -3.046289346362010

Which of the faster apps gives out garbage results?
SSE 2 or SSE3? Or all? *scratches head*

I have set MW to NNW until this gets straightened out.
____________

Detaching all boxes as empty. Permanently.

I have set my PC to NNW as well (will be updated the next time that PC connects to BAM).

I have set NW to NNW until this gets straightened out.

Faster applications returns incorrect results.

Original app:
fitness: -3.046272445362277

"Faster app":
fitness: -3.046289346362010

Faster applications returns incorrect results.

Original app:
fitness: -3.046272445362277

"Faster app":
fitness: -3.046289346362010

sesef, could you also check your faster app you are using please?

I just checked the results on the couple workunits crunched with this app and they are coming up as valid. Any ideas on specifically which app is causing an issue?

This is strange as well. I detached all boxes and the account page still shows all. Some even have work shown pending.

So how can we verify if the work being returned is correct, what should we look for?

Faster applications returns incorrect results.

Original app:
fitness: -3.046272445362277

"Faster app":
fitness: -3.046289346362010

sesef, could you also check your faster app you are using please?

which applications do you mean?

This is strange as well. I detached all boxes and the account page still shows all. Some even have work shown pending.

Thank you very much Cori!

So how can we verify if the work being returned is correct, what should we look for?

So how can we verify if the work being returned is correct, what should we look for?

The way I check is to go to my MW account, click on "Computers on this account" and then click individually on the tasks (under computer ID).

If you scroll to the next 20 results, and on, it will show whether the WU results are valid as the results given are quoted as successful under the title Outcome.

Faster applications returns incorrect results.

Original app:
fitness: -3.046272445362277

"Faster app":
fitness: -3.046289346362010

sesef, could you also check your faster app you are using please?

which applications do you mean?

sorry i thought in your team there where also optimized apps used made by some of your teammade.
i had the user mindc in my mind.
nevermind if im wrong.

correct me if i'm wrong

with
minimum quorum 1
initial replication 1
all results will be valid and give points, but it doesn't mean that they are correct :) to check this you have to compare results of the same wu crunched using opt app and official app
____________

correct me if i'm wrong

with
minimum quorum 1
initial replication 1
all results will be valid and give points, but it doesn't mean that they are correct :) to check this you have to compare results of the same wu crunched using opt app and official app

how do you crunch the same WU twice?

Travis has a post here where I presume the usual return of success is taken as a valid result as he will have knowledge of the outcome. Otherwise none of us can know if we are successfully churning out invalid results without a wingman to compare against.

I am aware the successfully returned results are being granted credit, as per the credit rules posted elsewhere.

But, can we assume the results are valid when it is shown as successful and been given credit?

So how can we verify if the work being returned is correct, what should we look for?

The way I check is to go to my MW account, click on "Computers on this account" and then click individually on the tasks (under computer ID).

If you scroll to the next 20 results, and on, it will show whether the WU results are valid as the results given are quoted as successful under the title Outcome.

Travis

I presume this can be done if someone was to D/L a batch of WUs and then back them up to another folder from the MW BOINC project folder.

If the WUs were then allowed to crunch in the normal way with this new faster client, with the no new taske active.

Then the backed up WUs were returned to the original folder and again crunched with the stock item.

Would this give an OK comparison (the results would be shown on your servers).

possible problem:
This means reporting to the MW project the same sets of results twice. Would that cause problems at your end?

So how can we verify if the work being returned is correct, what should we look for?

The way I check is to go to my MW account, click on "Computers on this account" and then click individually on the tasks (under computer ID).

If you scroll to the next 20 results, and on, it will show whether the WU results are valid as the results given are quoted as successful under the title Outcome.

This is not correct. WU's are being validated as correct, even though they may be returning incorrect data.

Beau

I do not think we can from our computers. As Travis posted here about the only method to verify that good results are being churned out.

Beau

I do not think we can from our computers. As Travis posted here about the only method to verify that good results are being churned out.

Beau

I do not think we can from our computers. As Travis posted here about the only method to verify that good results are being churned out.

And how would one do that? There were no sample inputs given with the code. Just the same link and app that others downloaded from arizmoon.com

Yes there are, if you go to http://milkyway.cs.rpi.edu/milkyway/download/code_release/
all of these have sample input files.

Beau

I do not think we can from our computers. As Travis posted here about the only method to verify that good results are being churned out.

And how would one do that? There were no sample inputs given with the code. Just the same link and app that others downloaded from arizmoon.com

Yes there are, if you go to http://milkyway.cs.rpi.edu/milkyway/download/code_release/
all of these have sample input files.

Ok, I downloaded the Milkyway release .7 (I am assuming that I was supposed to download the latest one?) Now what do I do with them? Sorry--I dont mean to be a pain...

Thanks

You should be able to put the milkyway executable and the other executable in the same directory with the sample inputs and run them. This would be (from where you uncompressed the files) ./milkyway/bin/test_files/

with the new changes over the weekend we will be not awarding credits to invalid results returned (and perhaps even docking them). We're still working out a good policy for this, so expect a news post soon.

Beau

I do not think we can from our computers. As Travis posted here about the only method to verify that good results are being churned out.

And how would one do that? There were no sample inputs given with the code. Just the same link and app that others downloaded from arizmoon.com

Yes there are, if you go to http://milkyway.cs.rpi.edu/milkyway/download/code_release/
all of these have sample input files.

Ok, I downloaded the Milkyway release .7 (I am assuming that I was supposed to download the latest one?) Now what do I do with them? Sorry--I dont mean to be a pain...

Thanks

You should be able to put the milkyway executable and the other executable in the same directory with the sample inputs and run them. This would be (from where you uncompressed the files) ./milkyway/bin/test_files/

Ok, copied them into the same folder and ran them, now what am I looking at?

I will put zip alternatives on www.arizmoon.com for the rar downloads later this evening.

Beau

I do not think we can from our computers. As Travis posted here about the only method to verify that good results are being churned out.

And how would one do that? There were no sample inputs given with the code. Just the same link and app that others downloaded from arizmoon.com

Yes there are, if you go to http://milkyway.cs.rpi.edu/milkyway/download/code_release/
all of these have sample input files.

Ok, I downloaded the Milkyway release .7 (I am assuming that I was supposed to download the latest one?) Now what do I do with them? Sorry--I dont mean to be a pain...

Thanks

You should be able to put the milkyway executable and the other executable in the same directory with the sample inputs and run them. This would be (from where you uncompressed the files) ./milkyway/bin/test_files/

Ok, copied them into the same folder and ran them, now what am I looking at?

in this folder you'll find file "out", last line of this file is your result.

If you scroll to the next 20 results, and on, it will show whether the WU results are valid as the results given are quoted as successful under the title Outcome.

If you scroll to the next 20 results, and on, it will show whether the WU results are valid as the results given are quoted as successful under the title Outcome.

You'd have to run the stock app stand-alone and check the result to see if the result of the computation is OK.

You should be able to put the milkyway executable and the other executable in the same directory with the sample inputs and run them. This would be (from where you uncompressed the files) ./milkyway/bin/test_files/

If you scroll to the next 20 results, and on, it will show whether the WU results are valid as the results given are quoted as successful under the title Outcome.

You'd have to run the stock app stand-alone and check the result to see if the result of the computation is OK.

Yes, and how would one do that? Where is the stock app executable located on the system?

If you scroll to the next 20 results, and on, it will show whether the WU results are valid as the results given are quoted as successful under the title Outcome.

You'd have to run the stock app stand-alone and check the result to see if the result of the computation is OK.

Yes, and how would one do that? Where is the stock app executable located on the system?

Okay.
Run BOINC for a bit on a MILKYWAY workunit.

Go here and scroll down to find the app suitable for your system (probably milkyway_07_windows_intel86 or x86_64) and download it.

Shut BOINC down, go to \Application Data\BOINC\Slots and create a new folder, paste the app. into it.

Look in the other slots and find one thats got a MILKYWAY part-finished in it, COPY files astronomy_parameters, search_parameters and stars to your new folder.

Doubleclick Milkyway_07 to start it up and sit till its finished running.

CUT your NEW FOLDER from there and PASTE it somewhere else (My Docs, perhaps) so as not to confuse BOINC when you start it up.

Now restart BOINC and let it finish the same work, set BOINC Network Activity Suspended so it wont upload the work as soon as its finished.

Open the RESULT you did with the new app (in My Docs\New Folder) with Wordpad, compare it with the result you'll find in Application Data\BOINC\projects\milkyway

Your standard app can typically be found on a Windows PC in

C:\Program Files\BOINC\projects\milkyway.cs.rpi.edu_milkyway\milkyway_0.7_windows_intelx86.exe
____________

Your standard app can typically be found on a Windows PC in

C:\Program Files\BOINC\projects\milkyway.cs.rpi.edu_milkyway\milkyway_0.7_windows_intelx86.exe

Any comfirmation if the new app's are accurate or not? I would like to keep using the sse2.rar version, but only if it is ok.
____________
Doesn't expecting the unexpected make the unexpected the expected?
If it makes sense, DON'T do it.

Your standard app can typically be found on a Windows PC in

C:\Program Files\BOINC\projects\milkyway.cs.rpi.edu_milkyway\milkyway_0.7_windows_intelx86.exe

There is no c:\program files\boinc\projects folder. If I look in the \all users\application data\boinc\projects\milkyway.cs.rpi.edu.milkyway folder; there is no milkyway_0.7_windows_intelx86.exe file in there at all (not even anything close to it) I have been running the stock app since I started with milkyway and I did not overwrite any files when copying the custom app in.

Any comfirmation if the new app's are accurate or not? I would like to keep using the sse2.rar version, but only if it is ok.

Your standard app can typically be found on a Windows PC in

C:\Program Files\BOINC\projects\milkyway.cs.rpi.edu_milkyway\milkyway_0.7_windows_intelx86.exe

There is no c:\program files\boinc\projects folder. If I look in the \all users\application data\boinc\projects\milkyway.cs.rpi.edu.milkyway folder; there is no milkyway_0.7_windows_intelx86.exe file in there at all (not even anything close to it) I have been running the stock app since I started with milkyway and I did not overwrite any files when copying the custom app in.

It (Boinc) will be wherever you installed it to. Or search for Boinc. Then look in that.

You usually have to use 'C:\Documents and Settings\All Users\Application Data\BOINC' for XP

or 'C:\ProgramData\BOINC' for Vista
____________
~It only takes one bottle cap moving at 23,000 mph to ruin your whole day~

You usually have to use C:\Documents and Settings\All Users\Application Data\BOINC for XP

It differs based on the OS. See above for one of the Vista pathways (there are multiple). Also Vista users beware that you have to enable hidden files to see the ProgramData file under your C drive, and that sometimes BOINC installs itself under the Users file instead.
____________
~It only takes one bottle cap moving at 23,000 mph to ruin your whole day~

Your standard app can typically be found on a Windows PC in

C:\Program Files\BOINC\projects\milkyway.cs.rpi.edu_milkyway\milkyway_0.7_windows_intelx86.exe

There is no c:\program files\boinc\projects folder. If I look in the \all users\application data\boinc\projects\milkyway.cs.rpi.edu.milkyway folder; there is no milkyway_0.7_windows_intelx86.exe file in there at all (not even anything close to it) I have been running the stock app since I started with milkyway and I did not overwrite any files when copying the custom app in.

Your standard app can typically be found on a Windows PC in

C:\Program Files\BOINC\projects\milkyway.cs.rpi.edu_milkyway\milkyway_0.7_windows_intelx86.exe

There is no c:\program files\boinc\projects folder. If I look in the \all users\application data\boinc\projects\milkyway.cs.rpi.edu.milkyway folder; there is no milkyway_0.7_windows_intelx86.exe file in there at all (not even anything close to it) I have been running the stock app since I started with milkyway and I did not overwrite any files when copying the custom app in.

Becouse you are using the opti app. If you rename the app_info.xml file, then BOINC downloads the stock app in that folder. (after stop and restart BOINC, of course)

Boinc deletes the standard app whenever you add an optimized one to the folder.
____________
~It only takes one bottle cap moving at 23,000 mph to ruin your whole day~

Your standard app can typically be found on a Windows PC in

C:\Program Files\BOINC\projects\milkyway.cs.rpi.edu_milkyway\milkyway_0.7_windows_intelx86.exe

There is no c:\program files\boinc\projects folder. If I look in the \all users\application data\boinc\projects\milkyway.cs.rpi.edu.milkyway folder; there is no milkyway_0.7_windows_intelx86.exe file in there at all (not even anything close to it) I have been running the stock app since I started with milkyway and I did not overwrite any files when copying the custom app in.

Becouse you are using the opti app. If you rename the app_info.xml file, then BOINC downloads the stock app in that folder. (after stop and restart BOINC, of course)

Ok; but I was using the standard app yesterday; and didnt delete anything. Would boinc have deleted it by itself?

Your standard app can typically be found on a Windows PC in

C:\Program Files\BOINC\projects\milkyway.cs.rpi.edu_milkyway\milkyway_0.7_windows_intelx86.exe

There is no c:\program files\boinc\projects folder. If I look in the \all users\application data\boinc\projects\milkyway.cs.rpi.edu.milkyway folder; there is no milkyway_0.7_windows_intelx86.exe file in there at all (not even anything close to it) I have been running the stock app since I started with milkyway and I did not overwrite any files when copying the custom app in.

Becouse you are using the opti app. If you rename the app_info.xml file, then BOINC downloads the stock app in that folder. (after stop and restart BOINC, of course)

Ok; but I was using the standard app yesterday; and didnt delete anything. Would boinc have deleted it by itself?

Yes. Is the standar actuation of BOINC.

Your standard app can typically be found on a Windows PC in

C:\Program Files\BOINC\projects\milkyway.cs.rpi.edu_milkyway\milkyway_0.7_windows_intelx86.exe

There is no c:\program files\boinc\projects folder. If I look in the \all users\application data\boinc\projects\milkyway.cs.rpi.edu.milkyway folder; there is no milkyway_0.7_windows_intelx86.exe file in there at all (not even anything close to it) I have been running the stock app since I started with milkyway and I did not overwrite any files when copying the custom app in.

Becouse you are using the opti app. If you rename the app_info.xml file, then BOINC downloads the stock app in that folder. (after stop and restart BOINC, of course)

Ok; but I was using the standard app yesterday; and didnt delete anything. Would boinc have deleted it by itself?

Yes. Is the standar actuation of BOINC.

Ahhh, well that would explain why I couldnt find it then, lol!! ok, so then I will remove the opti app, let boinc copy down the standard app, then copy the executable for it and ru the test to see if I get the same result. I will let you know what happens!

Your standard app can typically be found on a Windows PC in

C:\Program Files\BOINC\projects\milkyway.cs.rpi.edu_milkyway\milkyway_0.7_windows_intelx86.exe

There is no c:\program files\boinc\projects folder. If I look in the \all users\application data\boinc\projects\milkyway.cs.rpi.edu.milkyway folder; there is no milkyway_0.7_windows_intelx86.exe file in there at all (not even anything close to it) I have been running the stock app since I started with milkyway and I did not overwrite any files when copying the custom app in.

Becouse you are using the opti app. If you rename the app_info.xml file, then BOINC downloads the stock app in that folder. (after stop and restart BOINC, of course)

Ok; but I was using the standard app yesterday; and didnt delete anything. Would boinc have deleted it by itself?

Yes. Is the standar actuation of BOINC.

Ahhh, well that would explain why I couldnt find it then, lol!! ok, so then I will remove the opti app, let boinc copy down the standard app, then copy the executable for it and ru the test to see if I get the same result. I will let you know what happens!

Don't remove the opti app only rename the app_info.xml. If you remove the opti, BOINC will be return your actual wus with error when it can't find an app with to crunch them. When the next wu is downloaded, then the stock app too.

I am having problems doing the comparison. I downloaded the test files without any problem, then duplicated the folder, one to run with the stock app, and one to run with the opt app. Oddly enough, I was able to run the opt app in stand alone mode just fine. It is with the stock app that I am having problems.

The opt app (astronomy_sse3_app) "out" file says:

searchname
parameters [8]: 0.670000000000000 13.500000000000000 -1.734601060000000 33.352644669999997 27.423781280000000 1.244802812944947 1.914950983451476 2.500000000000000
metadata: this is the metadata
fitness: -3.106718478181371



But when I try to run the stock app (milkyway_0.7_windows_x86_64) in stand alone mode, I get no "out" file. The stderr file says:

Will run in standalone mode.
APP: error resolving search parameters file (for write): -108
filename: out
resolved output path: out

So I've hit a wall, in trying to compare the stock app to the opt app. FWIW, this is with XP64.
____________

You might need to remove the previous "out" file (generated by the other app) and the checkpoint file it generates.

@ Travis

Why linux, OSX, and Win stock apps gave different results with the same wu?

Best regards.
____________
Logan.

BOINC FAQ Service (Ahora, también disponible en Español/Now available in Spanish)

you might need to just put an empty "out" file in the directory for the stock app... i think i've seen this problem a couple times.

@ Travis

Why linux, OSX, and Win stock apps gave different results with the same wu?

Best regards.

At what level of accuracy?

has any one checked the sse2 out too see if its working right
____________

I just ran the SSE2 app on that same machine. The results matched the SSE3 results. So neither SSE2 or SSE3 match the stock app.
____________

I just ran the SSE2 app on that same machine. The results matched the SSE3 results. So neither SSE2 or SSE3 match the stock app.

So I guess this SSE3-Version saves the party for us.

Result of the Test-WU:
fitness: -3.046272445362275

I just ran the SSE2 app on that same machine. The results matched the SSE3 results. So neither SSE2 or SSE3 match the stock app.

Are these the apps from the MW.rar download?

If not, would you please specify exactly which apps you tested?

Thanks

Yes, they were both from the rar download.

So where do we stand?

As Labbie said can people who have tested specify what download they used. Also can anyone get in contact with the guys/gals (or they themselves) who made the apps and find out if they did any testing before release. For your information I have been using the SSE3 version from the MW.rar file downloaded from Ice's site.

For the moment I have set the machine to NNW and I will let it crunch out the remaining units. Then I guess I'll take it elsewhere while this is all sorted out.

And can anyone confirm what Logan brought up?

@ Travis

Why linux, OSX, and Win stock apps gave different results with the same wu?

So I guess this SSE3-Version saves the party for us.

Result of the Test-WU:
fitness: -3.046272445362275

Yes, I also got
fitness: -3.046272445362275

But that still does not match the stock app output of
fitness: -3.046272445362277

And can anyone confirm what Logan brought up?

@ Travis

Why linux, OSX, and Win stock apps gave different results with the same wu?

If this is the case does that not mean there could be an issue with the stock app too?

At what level of accuracy?

And can anyone confirm what Logan brought up?

@ Travis

Why linux, OSX, and Win stock apps gave different results with the same wu?

If this is the case does that not mean there could be an issue with the stock app too?

There are some issues with the stock app, but this accuracy thing is nothing you can really avoid between different architectures when using standard math libraries. And Travis answer (actually a question) to that problem

At what level of accuracy?

implies he is aware of that.

Actually, this sort of differences in the last bits is what one often calls numerical noise. Both versions are likely to calculate results within the same error bars when compared to a (hypothetical) implementation with infinite (or just somewhat extended) precision.

It's like Einstein and akosf all over again. Good times.
____________

Yeah that's fine. I think the cutoff where it would start to matter would be differences greater than 10^-10ish

Could someone please post a link to which one of these apps are supposedly ok. It's kind of confusing as there is more than one file download linked in the thread.

thanks.

Could someone please post a link to which one of these apps are supposedly ok. It's kind of confusing as there is more than one file download linked in the thread.

thanks.

However there is more than one sse3 app linked in this thread

one is

astronomy_sse3_app and is 889kb


and the other is


astronomy_SSE3_s and is 920kb

So I guess this SSE3-Version saves the party for us.

Result of the Test-WU:
fitness: -3.046272445362275

Yes, I also got
fitness: -3.046272445362275

But that still does not match the stock app output of
fitness: -3.046272445362277

Are they close enough it won't change the outcome (of the search) or should they be exact?

Yeah that's fine. I think the cutoff where it would start to matter would be differences greater than 10^-10ish

Ok so that would make it the


astronomy_SSE3_s one


cool cause thats the one I'm using

So, we have a new version of the SSE3 app that meets the 10 digit requirement. Does anyone have an updated SSE or SSE2 version?
____________

@ Travis

Why linux, OSX, and Win stock apps gave different results with the same wu?

Best regards.

At what level of accuracy?

Travis/Dave

Part of the assimilator update allows us to track users that repeatedly return invalid results and stop awarding credit to them

Travis/Dave

Part of the assimilator update allows us to track users that repeatedly return invalid results and stop awarding credit to them

How quickly do you guys expect this assimilator update to pick up users returning invalid results and stop awarding credit?

Is it very soon (<1 day) or more?

So, we have a new version of the SSE3 app that meets the 10 digit requirement. Does anyone have an updated SSE or SSE2 version?

Heres what I got for the SSEv2 zip

Optimized
fitness: -3.046272445362275

Stock
fitness: -3.046272445361065


So are these results too different?

Heres what I got for the SSEv2 zip

Optimized
fitness: -3.046272445362275

Stock
fitness: -3.046272445361065


So are these results too different?

So, we have a new version of the SSE3 app that meets the 10 digit requirement. Does anyone have an updated SSE or SSE2 version?

Here we go...

http://rapidshare.com/files/184217251/astronomy_sse2_app.rar.html

Original app:
fitness: -3.046272445362277

this one:
fitness: -3.046272445362278

have fun...

LB

too slow....

I don t think so.
This is what Travis wrote:
Yeah that's fine. I think the cutoff where it would start to matter would be differences greater than 10^-10ish

So the last five digits are not necessary,i think...
There are only the last four incorrect...

Correct me,if i am wrong...

LB

Gonna try re-attaching with one SSE3 box. If I end up losing points I will shut it back down.

Gonna try re-attaching with one SSE3 box. If I end up losing points I will shut it back down.

Yeah this is fine. It would be nice to see how it does with a couple different inputs, just to make sure. Is this optimized app done with compiler flags or has there been significant code changes? If the 2nd is the case it would be nice if they were shared so we could update the stock app.

So I guess this SSE3-Version saves the party for us.

Result of the Test-WU:
fitness: -3.046272445362275

Yes, I also got
fitness: -3.046272445362275

But that still does not match the stock app output of
fitness: -3.046272445362277

Are they close enough it won't change the outcome (of the search) or should they be exact?

Yeah that's fine. I think the cutoff where it would start to matter would be differences greater than 10^-10ish

The app linked above is the only one I know of that is good. The last post above (the "Yeah that's fine" one) is from Travis.

Gonna try re-attaching with one SSE3 box. If I end up losing points I will shut it back down.

I've been running with the latest SSE3 app all day without problems. The WUs are taking slightly (minimally) longer, but it's still turning WUs around far faster than the stock app.

Pentium D & P-4 HT are validating. Quad core AMD Phenom now attached.

Heres what I got for the SSEv2 zip

Optimized
fitness: -3.046272445362275

Stock
fitness: -3.046272445361065


So are these results too different?

Would be interesting to know if the app still suffers from the memleak that was 'build in' to the stock app and what's more important, if it still cuts off significant numbers when restated form a checkpoint like the stock app does (which of course makes the result invalid )

Heres what I got for the SSEv2 zip

Optimized
fitness: -3.046272445362275

Stock
fitness: -3.046272445361065


So are these results too different?

Would be interesting to know if the app still suffers from the memleak that was 'build in' to the stock app and what's more important, if it still cuts off significant numbers when restated form a checkpoint like the stock app does (which of course makes the result invalid )

At least my SSE3-App I linked to in this thread does not suffer from any of these shortcomings ;)

I'm pretty sure the newest versions of the stock app checkpoint to 20 digits past the decimal point so that isn't happening anymore.

*edit* And honestly, if there are bugs in the application that you know about, please post them to the code discussion forum. If you make a post about them there (instead of random snarky comments in random threads), it would actually be helpful to us so we can update the application and get rid of them.

At least my SSE3-App I linked to in this thread does not suffer from any of these shortcomings ;)

I for one am grateful for you making this available Cluster Physik.

I've already added it to www.arizmoon.com which I will be updating now with what I think are the 'correct' SSE2 and SSE apps.