Which patch did you use?

The one Simon provided seems to work, I've got a couple of valid results:

http://setiathome.berkeley.edu/result.php?resultid=471542706
http://setiathome.berkeley.edu/result.php?resultid=471542682

These two long (62.40 cr) units seemed a little bit faster (maybe a 10%), yet it's too early to say anything for sure.

---

Looking good - like I said, model-dependent, it's either quicker or slower. Haven't seen one where patched and unpatched are level, so far.

<OT plug>
Quick question: Have you donated lately?
SETI@Home needs your help, and now you can donate via PayPal, too!

Every little bit helps.
</OT plug>

Regards,
Simon.

---

Donate to SETI@Home via PayPal!

Optimized SETI@Home apps + Information

Looking good - like I said, model-dependent, it's either quicker or slower. Haven't seen one where patched and unpatched are level, so far.

Regards,
Simon.

I'm trying it on my Athalon 64 X2 4400+. I think it will be hard to tell if it's 10% better or 10% worse since, for various reasons, my RAC is climbing from a lower level. I'm also OC'd from 2.2GHz to 2.6GHz on air. Hoping to eventually hit about a RAC of about 1100 . . . maybe when it seems that I've leveled off, I can put the generec SSE2 back in place and see if my RAC drops . . .

---

Looking good - like I said, model-dependent, it's either quicker or slower. Haven't seen one where patched and unpatched are level, so far.

<OT plug>
Quick question: Have you donated lately?
SETI@Home needs your help, and now you can donate via PayPal, too!

Every little bit helps.
</OT plug>

Regards,
Simon.

I will let it run for a few days as right now with the units we are getting it is hard to tell if I am faster or not...Glad to see my Opty can run the windows version...
<core_client_version>5.9.0</core_client_version>
<![CDATA[
<stderr_txt>
Optimized SETI@Home Enhanced application

Optimizers: Ben Herndon, Josef Segur, Alex Kan, Simon Zadra
Version: Windows SSE3 32-bit 'Ni!' based on seti V5.15 'Chicken Good!'
Rev: (R-2.0|QxP|FFT:IPP_SSE3|Ben-Joe)
CPUID: 'AMD K8 Opteron DC 1 (Denmark)'
cpus: 1 cores: 2 threads: 1 cache: L1=64K L2=1024K L3=0K
features: mmx 3Dnow 3Dnow+ sse sse2 sse3
speed: 2574 MHz -- read megs/sec: L1=14113, L2=7003, RAM=3390

TOO BAD NONE ARE 64-BIT

---

Official Abuser of Boinc Buttons...
And no good credit hound!

I used Simons patch and an SSE3 Intel type Seti app from His website. The minute I launched Boinc with It I got computation errors, I don't think 5.9.0.32(by Crunch3r) had anything to do with It, But like I said I'd made a backup and restored It very, very quickly.

@Batman-

I also got some computational errors. If I remember correctly, it was because there was a mis-match between the new application name and the app name in the app_info.xml file (or at least that was one of my problems). Once I had that fixed, it downloaded new WUs and has been crunching happily since. Unfortunately I wasn't able to recover the comp error WUs . . .

HTH,
Brock

---

I will let it run for a few days as right now with the units we are getting it is hard to tell if I am faster or not...Glad to see my Opty can run the windows version...
<core_client_version>5.9.0</core_client_version>
<![CDATA[
<stderr_txt>
Optimized SETI@Home Enhanced application

Optimizers: Ben Herndon, Josef Segur, Alex Kan, Simon Zadra
Version: Windows SSE3 32-bit 'Ni!' based on seti V5.15 'Chicken Good!'
Rev: (R-2.0|QxP|FFT:IPP_SSE3|Ben-Joe)
CPUID: 'AMD K8 Opteron DC 1 (Denmark)'
cpus: 1 cores: 2 threads: 1 cache: L1=64K L2=1024K L3=0K
features: mmx 3Dnow 3Dnow+ sse sse2 sse3
speed: 2574 MHz -- read megs/sec: L1=14113, L2=7003, RAM=3390

TOO BAD NONE ARE 64-BIT

@hiamps-
Here's my WU data-

<core_client_version>5.8.8</core_client_version>
<![CDATA[
<stderr_txt>
Optimized SETI@Home Enhanced application

Optimizers: Ben Herndon, Josef Segur, Alex Kan, Simon Zadra
Version: Windows SSE2 32-bit 'Ni!' based on seti V5.15 'Chicken Good!'
Rev: (R-2.0|QxN|FFT:IPP_SSE2|Ben-Joe)
CPUID: 'AMD K8 Athlon 64 X2 (Toledo)'
cpus: 1 cores: 2 threads: 1 cache: L1=64K L2=1024K L3=0K
features: mmx 3Dnow 3Dnow+ sse sse2 sse3
speed: 2597 MHz -- read megs/sec: L1=14254, L2=7299, RAM=3185

Work Unit Info
True angle range: 1.582064

Spikes Pulses Triplets Gaussians Flops
2 0 0 0 3840658301454
</stderr_txt>
]]>

---

Brock which of these are You using with iccpatch.exe?

SSE2-Intel P-M optimized Windows S@H app
or
SSE2-Intel P4 optimized Windows S@H app

Hey Bats-

I've patched and am using the SSE2-Intel P4 Chicken Soup Science App.

HTH,
Brock

---

I tried a patched version that I made using the Simon patch again this time the SSE2 version(first I did this, I patched the file under XP x64, 2nd time under XP Pro x32) and I still got computation errors no matter what, I made sure both files were alone this time in a folder on the desktop. when I type in iccpatch.exe sah_5.15.exe(I shortened the name this time too and renamed It afterwards) It went like It's supposed to and It patched in 10 places according to the program. Am I doing something wrong?

I followed simons instructions and the only problem I had was figuring out to add the .exe to my client as it didn't say it in the folder I had it. Then it all went smooth. I will try the intel SSE3 for a few then will patch the SSE2 and see which is faster for my machine. I am using XP 64, works great on my machine and I have no clue how to use DOS...Had a heck of a time getting to just a c: prompt mine opened in Documents and settings....

---

Official Abuser of Boinc Buttons...
And no good credit hound!

I will let it run for a few days as right now with the units we are getting it is hard to tell if I am faster or not...Glad to see my Opty can run the windows version...
<core_client_version>5.9.0</core_client_version>
<![CDATA[
<stderr_txt>
Optimized SETI@Home Enhanced application

Optimizers: Ben Herndon, Josef Segur, Alex Kan, Simon Zadra
Version: Windows SSE3 32-bit 'Ni!' based on seti V5.15 'Chicken Good!'
Rev: (R-2.0|QxP|FFT:IPP_SSE3|Ben-Joe)
CPUID: 'AMD K8 Opteron DC 1 (Denmark)'
cpus: 1 cores: 2 threads: 1 cache: L1=64K L2=1024K L3=0K
features: mmx 3Dnow 3Dnow+ sse sse2 sse3
speed: 2574 MHz -- read megs/sec: L1=14113, L2=7003, RAM=3390

TOO BAD NONE ARE 64-BIT

@hiamps-
Here's my WU data-

<core_client_version>5.8.8</core_client_version>
<![CDATA[
<stderr_txt>
Optimized SETI@Home Enhanced application

Optimizers: Ben Herndon, Josef Segur, Alex Kan, Simon Zadra
Version: Windows SSE2 32-bit 'Ni!' based on seti V5.15 'Chicken Good!'
Rev: (R-2.0|QxN|FFT:IPP_SSE2|Ben-Joe)
CPUID: 'AMD K8 Athlon 64 X2 (Toledo)'
cpus: 1 cores: 2 threads: 1 cache: L1=64K L2=1024K L3=0K
features: mmx 3Dnow 3Dnow+ sse sse2 sse3
speed: 2597 MHz -- read megs/sec: L1=14254, L2=7299, RAM=3185

Work Unit Info
True angle range: 1.582064

Spikes Pulses Triplets Gaussians Flops
2 0 0 0 3840658301454
</stderr_txt>
]]>

Wonder why your times are nowhere near mine? Our machines seem quite close. Mine will do 62's in the 7000 second range, even with the generic...Check out my opty times...

---

Official Abuser of Boinc Buttons...
And no good credit hound!

@Bats-

One last thing-

I used the executable version and it gives you the error message about '10 instances and expecting a multiple of 3' but it still seems to have made the approprate changes as opposed to the strictly Perl script which skipped the mods when I tried it. Don't forget to make sure the app filename is changed in all the appropriate places in the app_info.xml file. Here's the steps that worked for me-

- Stopped BOINC service.
- Backed up BOINC folder.
- Copied patched SSE2-Intel science app into the SaH project folder.
- Modified the app_info.xml file in the SaH project folder with the new apps name.
- Restarted BOINC service.
- Sat back and started watching my RAC . . .

HTH,
Brock

---

Wonder why your times are nowhere near mine? Our machines seem quite close. Mine will do 62's in the 7000 second range, even with the generic...Check out my opty times...

@hiamps-

You're right. Must be something internal to the Opteron vs. the Athalon X2. I was doing about 11,000 seconds at a stock 2.2 GHz and now I'm taking about 9,500 seconds at an OC'd 2.6 GHz. Even our L1, L2, and RAM specs are almost identical . . .

Maybe another cruncher with more AMD CPU knowledge could shed some light on this . . .

Brock

---

Wonder why your times are nowhere near mine? Our machines seem quite close. Mine will do 62's in the 7000 second range, even with the generic...Check out my opty times...

@hiamps-

You're right. Must be something internal to the Opteron vs. the Athalon X2. I was doing about 11,000 seconds at a stock 2.2 GHz and now I'm taking about 9,500 seconds at an OC'd 2.6 GHz. Even our L1, L2, and RAM specs are almost identical . . .

Maybe another cruncher with more AMD CPU knowledge could shed some light on this . . .

Brock

Brock,
If you are using an ASUS board you have to manually set memory to 1T as it is default 2 T. With your overclock you probably know more about that than I but a friend didn't so I figured I would add that. I can't get past where I am with no errors, keep threatening to break out my water cooling just to see what this will do.

---

Official Abuser of Boinc Buttons...
And no good credit hound!

Wonder why your times are nowhere near mine? Our machines seem quite close. Mine will do 62's in the 7000 second range, even with the generic...Check out my opty times...

@hiamps-

You're right. Must be something internal to the Opteron vs. the Athalon X2. I was doing about 11,000 seconds at a stock 2.2 GHz and now I'm taking about 9,500 seconds at an OC'd 2.6 GHz. Even our L1, L2, and RAM specs are almost identical . . .

Maybe another cruncher with more AMD CPU knowledge could shed some light on this . . .

Brock

Will give this a try...I put a pic of my CPUZ screen

hiamps CPUZ

---

Official Abuser of Boinc Buttons...
And no good credit hound!

You're right. Must be something internal to the Opteron vs. the Athalon X2.

The Dual Core Opteron 1xx has 1 MB L2 cache per core, while your 4400+ should have only 512 kB.

But I don't think it should make such a big difference, my Athlon64 3200+ Winchester (single core with 512 kB L2) did the 62.40 in 8500-8900 seconds at 2300 MHz with KWSN 2.0 SSE2-Generic.

---

@Bats-

One last thing-

I used the executable version and it gives you the error message about '10 instances and expecting a multiple of 3' but it still seems to have made the approprate changes as opposed to the strictly Perl script which skipped the mods when I tried it. Don't forget to make sure the app filename is changed in all the appropriate places in the app_info.xml file. Here's the steps that worked for me-

- Stopped BOINC service.
- Backed up BOINC folder.
- Copied patched SSE2-Intel science app into the SaH project folder.
- Modified the app_info.xml file in the SaH project folder with the new apps name.
- Restarted BOINC service.
- Sat back and started watching my RAC . . .

HTH,
Brock

- I stopped Boinc
- I backed Boinc up
- copied the patched SSE2-Intel app into the sah folder
- I used the app_info.xml that came with the intel app, Not the one I'd been using with the generic app.
- restarted boinc
- sat and watched a few computational errors and quit Boinc
- restored last Boinc install.

I copied the .exe, the .pdb and the application into my boinc/projects. I also found those same old files in there and deleted them before I restarted. And I saved a copy of Boinc as Boinc1 just in case...So far so good. Hope this helps. And during the patch it did find 10 instances...

---

Official Abuser of Boinc Buttons...
And no good credit hound!

Will give this a try...I put a pic of my CPUZ screen

hiamps CPUZ

@hiamp-

Thanks for the tip. My CPUZ info was nearly identical to yours except for the 1T/2T setting. I've set it from 2T to 1T. Let's sit back and see what happens . . .

Brock

---

@Bats-

One last thing-

I used the executable version and it gives you the error message about '10 instances and expecting a multiple of 3' but it still seems to have made the approprate changes as opposed to the strictly Perl script which skipped the mods when I tried it. Don't forget to make sure the app filename is changed in all the appropriate places in the app_info.xml file. Here's the steps that worked for me-

- Stopped BOINC service.
- Backed up BOINC folder.
- Copied patched SSE2-Intel science app into the SaH project folder.
- Modified the app_info.xml file in the SaH project folder with the new apps name.
- Restarted BOINC service.
- Sat back and started watching my RAC . . .

HTH,
Brock

- I stopped Boinc
- I backed Boinc up
- copied the patched SSE2-Intel app into the sah folder
- I used the app_info.xml that came with the intel app, Not the one I'd been using with the generic app.
- restarted boinc
- sat and watched a few computational errors and quit Boinc
- restored last Boinc install.

I copied the .exe, the .pdb and the application into my boinc/projects. I also found those same old files in there and deleted them before I restarted. And I saved a copy of Boinc as Boinc1 just in case...So far so good. Hope this helps. And during the patch it did find 10 instances...

Nope, I did what You did, twice, I used My old app info file after making sure everything was correct by doing a copy and paste so as to be no mispellings, then I took the name "-intel" out minus the quotes of course, I'm batting a thousand. Is anyone using Crunch3r's Boinc 5.9.0.32 client? Otherwise I'm out of ideas, I used the iccpatch.exe of course, So I'm stumped. And My Opterons are crunching as usual.

I didn't do anything to the old app info file or any "-intel" What is that? The units that were going finished as the old app then the new ones say SSE3.

---

Official Abuser of Boinc Buttons...
And no good credit hound!

I copied the .exe, the .pdb and the application into my boinc/projects. I also found those same old files in there and deleted them before I restarted. And I saved a copy of Boinc as Boinc1 just in case...So far so good. Hope this helps. And during the patch it did find 10 instances...

Nope, I did what You did, twice, I used My old app info file after making sure everything was correct by doing a copy and paste so as to be no mispellings, then I took the name "-intel" out minus the quotes of course, I'm batting a thousand. Is anyone using Crunch3r's Boinc 5.9.0.32 client? Otherwise I'm out of ideas, I used the iccpatch.exe of course, So I'm stumped. And My Opterons are crunching as usual.[/quote]

@Bats-

When I got my computational errors (all the WUs in my queue), I ended up hitting the "Update Project" button and it uploaded the comp error WUs and then downloaded a new set of WUs and that's what my machine is working on now. I don't know if you can make this change without getting comp errors on your existing set of queued-up WUs or not. That was my experience. I would have rather not "comp errored" all my queued WUs but that's the way it worked out.

HTH,
Brock

---

I copied the .exe, the .pdb and the application into my boinc/projects. I also found those same old files in there and deleted them before I restarted. And I saved a copy of Boinc as Boinc1 just in case...So far so good. Hope this helps. And during the patch it did find 10 instances...

Nope, I did what You did, twice, I used My old app info file after making sure everything was correct by doing a copy and paste so as to be no mispellings, then I took the name "-intel" out minus the quotes of course, I'm batting a thousand. Is anyone using Crunch3r's Boinc 5.9.0.32 client? Otherwise I'm out of ideas, I used the iccpatch.exe of course, So I'm stumped. And My Opterons are crunching as usual.

@Bats-

When I got my computational errors (all the WUs in my queue), I ended up hitting the "Update Project" button and it uploaded the comp error WUs and then downloaded a new set of WUs and that's what my machine is working on now. I don't know if you can make this change without getting comp errors on your existing set of queued-up WUs or not. That was my experience. I would have rather not "comp errored" all my queued WUs but that's the way it worked out.

HTH,
Brock[/quote]

Hi Brock,
I didn't update anything and the old units were half done completed as usual. all the rest in my que have been going no problem.

---

Official Abuser of Boinc Buttons...
And no good credit hound!

One more thing...Even though I right clicked my Boinc icon and closed boinc I could not delete a icon that was in use...I went into task manager and there were 2 instances of CLi.exe running I had to maually kill then it worked.

---

Official Abuser of Boinc Buttons...
And no good credit hound!