Hi Folks,

Just got around to recompiling the latest 4.19 boinc client with optimizations for various AMD and Intel processors and have now posted them to my site for download.

There are AMD (K6-2, K6-3, Athlon and AthlonXP) and Intel (PIII, P4 and i686) versions available. They are all 32-bit, I believe Aki has the AMD 64-bit client if you're that way inclined ;)

I've not had a chance to thoroughly test them yet, so I'd really appreciate if anyone who uses them could post how you get on with them.

Get them here (or click on the link in my sig):

http://www.pperry.f2s.com/

Enjoy,

Ned

---

*** My Guide to Compiling Optimised BOINC and SETI Clients ***
*** Download Optimised BOINC and SETI Clients for Linux Here ***

Hi Ned

Thanks for the recompiled 4.19 boinc client, I'm running it on a P4.

4.13 client
1579 Whetstone
2487 Dhrystone

4.19 client
1625 Whetstone
2789 Dhrystone

A nice increase and no problems encountered.

Thanks again.

Don

---

Ned,

Thanks for the updated client. I'm running it on a P4 and SUSE Linux 9.1. It has been running for about 14 hrs. with no problems.

Bob

---

Thanks for the feedback guys :)

Anyone else using any of them besides the P4 client?

Ned

---

*** My Guide to Compiling Optimised BOINC and SETI Clients ***
*** Download Optimised BOINC and SETI Clients for Linux Here ***

> Anyone else using any of them besides the P4 client?

Hi Ned,

thanks a lot for the actual clients. I've just updated a few of my older stack nodes.

K6-2 400
4.13 public client
164 Whetstone
397 Dhrystone

4.19 optimised client
425 Whetstone
776 Dhrystone
-----------------------
P3 500
4.13 public client
264 Whetstone
513 Dhrystone

4.19 optimised client
501 Whetstone
1032 Dhrystone
-----------------------
P3 550
4.13 public client
289 Whetstone
561 Dhrystone

4.19 optimised client
548 Whetstone
1129 Dhrystone

All of them are running without an issue. Later on i will probably run a few more tests on newer hardware.
Good job!

greetings from germany
rattelschneck

---

Ned thanks so much for your work it makes a big diff on my XP Linux machine (as you well know) so thanks!

---

<img src="http://www.boincstats.com/stats/banner.php?cpid=b3c0c2639ea110901bd0970a1c22efcd">

> Thanks for the feedback guys :)
>
> Anyone else using any of them besides the P4 client?
>
> Ned

I tried with a P4 and there was a huge difference in benchmarks. Now I'm going to update my K6 processor to check if there is any improvement. By the way, in /proc/cpuinfo it says:

vendor_id : AuthenticAMD
cpu family : 5
model : 8
model name : AMD-K6(tm) 3D processor

Do anyone know if what version should I use?

Thanks in advance

Zerjillo

---

I've just tried both K6-2 and K6-3 and the i686 on my AMD k6 3D and none of them works (SIGILL: illegal instruction). Maybe it is a very old processor and there is no possibility of improvement?

Any feedback will be welcome

Zerjillo

---

>
> I tried with a P4 and there was a huge difference in benchmarks. Now I'm going
> to update my K6 processor to check if there is any improvement. By the way, in
> /proc/cpuinfo it says:
>
> vendor_id : AuthenticAMD
> cpu family : 5
> model : 8
> model name : AMD-K6(tm) 3D processor
>
> Do anyone know if what version should I use?
>
> Thanks in advance
>
> Zerjillo
>

You need a client optimized to run on an AMD K6. I initially only optimized versions for more common processors that people are using.

I'll make you a client and post it to my site :)

For reference, below is the full list of processor types that can be specified and optimized for.

Ned

-------------------------------------------------------------------

i386
Original Intel's i386 CPU.

i486
Intel's i486 CPU. (No scheduling is implemented for this chip.)

i586, pentium
Intel Pentium CPU with no MMX support.

pentium-mmx
Intel PentiumMMX CPU based on Pentium core with MMX instruction set support.

i686, pentiumpro
Intel PentiumPro CPU.

pentium2
Intel Pentium2 CPU based on PentiumPro core with MMX instruction set support.

pentium3, pentium3m
Intel Pentium3 CPU based on PentiumPro core with MMX and SSE instruction set support.

pentium-m
Low power version of Intel Pentium3 CPU with MMX, SSE and SSE2 instruction set support. Used by Centrino notebooks.

pentium4, pentium4m
Intel Pentium4 CPU with MMX, SSE and SSE2 instruction set support.

prescott
Improved version of Intel Pentium4 CPU with MMX, SSE, SSE2 and SSE3 instruction set support.

nocona
Improved version of Intel Pentium4 CPU with 64-bit extensions, MMX, SSE, SSE2 and SSE3 instruction set support.

k6
AMD K6 CPU with MMX instruction set support.

k6-2, k6-3
Improved versions of AMD K6 CPU with MMX and 3dNOW! instruction set support.

athlon, athlon-tbird
AMD Athlon CPU with MMX, 3dNOW!, enhanced 3dNOW! and SSE prefetch instructions support.

athlon-4, athlon-xp, athlon-mp
Improved AMD Athlon CPU with MMX, 3dNOW!, enhanced 3dNOW! and full SSE instruction set support.

k8, opteron, athlon64, athlon-fx
AMD K8 core based CPUs with x86-64 instruction set support. (This supersets MMX, SSE, SSE2, 3dNOW!, enhanced 3dNOW! and 64-bit instruction set extensions.)

winchip-c6
IDT Winchip C6 CPU, dealt in same way as i486 with additional MMX instruction set support.

winchip2
IDT Winchip2 CPU, dealt in same way as i486 with additional MMX and 3dNOW! instruction set support.

c3
Via C3 CPU with MMX and 3dNOW! instruction set support. (No scheduling is implemented for this chip.)

c3-2
Via C3-2 CPU with MMX and SSE instruction set support. (No scheduling is implemented for this chip.)

---

*** My Guide to Compiling Optimised BOINC and SETI Clients ***
*** Download Optimised BOINC and SETI Clients for Linux Here ***

Thanks a lot Ned. I'll look forward to the K6 versión.

What I'm wondering now if why would the processor say:

AMD-K6(tm) 3D processor

when it does not implement the 3DNOW! instruction set (otherwise the K6-2 or K6-3 version would have worked).

Thanks again

Zerjillo

---

You're welcome.

No promises, but I'll try and do it today and post it for you (hopefully I'll have time).

Anyway, I'll post back here when it's done.

Ned

---

*** My Guide to Compiling Optimised BOINC and SETI Clients ***
*** Download Optimised BOINC and SETI Clients for Linux Here ***

OK, K6 client is available.

MD5 check sum for the .tar.bz2 file is:

e1f6360a71fafc423735b97376dd4f48 boinc_4.19_k6-pc-linux-gnu.tar.bz2

Please let me know how you get on with this client.

Regards,

Ned

---

*** My Guide to Compiling Optimised BOINC and SETI Clients ***
*** Download Optimised BOINC and SETI Clients for Linux Here ***

Thanks for the new version, but as for the others it still doesn't work:

----------------------------------------------
# ./boinc_4.19 -run_cpu_benchmarks
2005-01-30 16:08:10 [---] Starting BOINC client version 4.19 for i686-pc-linux-gnu
2005-01-30 16:08:10 [SETI@home] Project prefs: no separate prefs for home; using your defaults
2005-01-30 16:08:10 [SETI@home] Host ID is 480275
2005-01-30 16:08:10 [---] General prefs: from unknown project http://climateprediction.net/ (last modified 2004-09-13 18:51:18)
2005-01-30 16:08:10 [---] General prefs: no separate prefs for home; using your defaults
SIGILL: illegal instruction
Exiting...
----------------------------------------------

It seems to fail when beggining the cpu benchmarks. With the P4 version (in a P4, of course) it works, so something must be happening with this computer. Any ideas?

Zerjillo

---

> SIGILL: illegal instruction
> Exiting...

Is the the exact error messages?

---

Hmm - strange. That's normally the error you get when running on the wrong type of processor (ie running the P4 client on a non-P4 processor)

Does the standard berkeley client run?

Ned

---

*** My Guide to Compiling Optimised BOINC and SETI Clients ***
*** Download Optimised BOINC and SETI Clients for Linux Here ***

The exact message is the one I posted (copy / paste)

And yes, the standard berkeley client runs perfectly.

Thanks again

Zerjillo

---

> The exact message is the one I posted (copy / paste)

Cool! I will add it to the messages in the FAQ. I just collected a couple new ones myself ... Or new in the sense I do not have them in the database yet.

---

Paul: Note that this is not a "BOINC" error message, but a linux one. It means that the programm contained one instruction that the processor does not understand, so the program aborts. This error should only happen when you use software that is compiled to work with a processor that has a different instruction set. That's the reason why much applications are compiled for the i386 instruction set: it is not very optimised but you "assure" that it will work on any i386 compatible processor.

Zerjillo

---

> Paul: Note that this is not a "BOINC" error message, but a linux one. It means
> that the programm contained one instruction that the processor does not
> understand, so the program aborts. This error should only happen when you use
> software that is compiled to work with a processor that has a different
> instruction set. That's the reason why much applications are compiled for the
> i386 instruction set: it is not very optimised but you "assure" that it will
> work on any i386 compatible processor.
>
> Zerjillo

Well, the average participant is not going to care about that ... :)

I agree that it is an OS trap message,but that is no reason to not cover it. Not that I have herds running to my site to look things up ... sigh ... :)

At least I do get SOME traffic. And I appreciate all of you that refer people to the site. It *IS* getting better. Heck I am, hmm, better not say this, migh jinx it ...

But I am making good progress on this pass through the documentation. With global edits and PHP functions I am getting a better consistancy on site appearance (if nothing else).

---

Correction: SIGILL applies to Unix, Linux, and BSD (OS X). It is not caused by BOINC, SETI, or any other project. If memory serves, the kernel is responsible for signal management.

---