YAOSCW r8.1 (optimized client for Windows)

Message boards : Number crunching : YAOSCW r8.1 (optimized client for Windows)
Message board moderation

To post messages, you must log in.

1 · 2 · 3 · 4 . . . 10 · Next

AuthorMessage
Tetsuji Maverick Rai
Volunteer tester
Avatar

Send message
Joined: 25 Apr 99
Posts: 518
Credit: 90,863
RAC: 0
Japan
Message 124157 - Posted: 16 Jun 2005, 10:08:20 UTC
Last modified: 16 Jun 2005, 10:17:35 UTC

I built new series using IPP4.1 and new Intel C++ compiler 9.0. Due to small source changes and version up of the compiler, it's slightly faster than r7 on my P4 Northwood and Prescott. app_info.xml and a client are included in each zip.

YAOSCW-P-r8.1.zip

YAOSCW-N-r8.1.zip

YAOSCW-B-r8.1.zip

YAOSCW-W-r8.1.zip

YAOSCW-K-r8.1.zip

YAOSCW-0-r8.1.zip

and the source is

YAOSCW-r8.1-src.zip

as for processor/name relationships see this page
Luckiest in the world. WMD = Weapon of Mass Distraction.
Click this table.
ID: 124157 · Report as offensive
Profile Paul D. Buck
Volunteer tester

Send message
Joined: 19 Jul 00
Posts: 3898
Credit: 1,158,042
RAC: 0
United States
Message 124177 - Posted: 16 Jun 2005, 11:09:55 UTC - in response to Message 124157.  


as for processor/name relationships see this page


I don't know about anyone else, but I see no relationship between the page you linked to and the file names ...
ID: 124177 · Report as offensive
Raithmir
Volunteer tester

Send message
Joined: 3 Apr 99
Posts: 89
Credit: 385,065
RAC: 0
United Kingdom
Message 124179 - Posted: 16 Jun 2005, 11:15:58 UTC - in response to Message 124177.  


as for processor/name relationships see this page


I don't know about anyone else, but I see no relationship between the page you linked to and the file names ...


The relevent part of the file name are the P, N, B, W, K, and 0 letters.
ID: 124179 · Report as offensive
Dusty33
Volunteer tester
Avatar

Send message
Joined: 16 Jul 01
Posts: 24
Credit: 3,368,100
RAC: 0
United States
Message 124181 - Posted: 16 Jun 2005, 11:24:15 UTC - in response to Message 124157.  

YAOSCW-P-r8.1.zip - QxP = P4 SSE3
YAOSCW-N-r8.1.zip - QxN = P4 SSE2 only
YAOSCW-B-r8.1.zip - QxB = Pentium M SSE2
YAOSCW-W-r8.1.zip - QxW = P4 SSE2 AMD64
YAOSCW-K-r8.1.zip - QxK = P3 SSE AMD-XP
YAOSCW-0-r8.1.zip - No Option = P2 MMX

Hope that helps.



Dusty
ID: 124181 · Report as offensive
Ned Slider

Send message
Joined: 12 Oct 01
Posts: 668
Credit: 4,375,315
RAC: 0
United Kingdom
Message 124204 - Posted: 16 Jun 2005, 13:28:52 UTC

I'm trying the K version now on an AMD Duron (morgan core) with SSE. It's running fine.

Ned


*** My Guide to Compiling Optimised BOINC and SETI Clients ***
*** Download Optimised BOINC and SETI Clients for Linux Here ***
ID: 124204 · Report as offensive
Profile Speedy67 & Friends
Volunteer tester
Avatar

Send message
Joined: 14 Jul 99
Posts: 335
Credit: 1,178,138
RAC: 0
Netherlands
Message 124231 - Posted: 16 Jun 2005, 15:19:49 UTC
Last modified: 16 Jun 2005, 15:21:54 UTC

I'm giving the P-version a try on my p4-prescott. Seems to be running fine. Thanks for another speedup. :)

Greetings,
Speedy67

[edit:] Of course also mirrored here. :)


ID: 124231 · Report as offensive
Divide Overflow
Volunteer tester
Avatar

Send message
Joined: 3 Apr 99
Posts: 365
Credit: 131,684
RAC: 0
United States
Message 124303 - Posted: 16 Jun 2005, 19:07:13 UTC
Last modified: 16 Jun 2005, 19:08:38 UTC

The v7 -B version was working wonderfully on my 1.8 Pentium M (Dothan).
I've updated to the new r8.1 and am crunching away now.

Thanks again, Tetsuji!!

ID: 124303 · Report as offensive
j2satx
Volunteer tester

Send message
Joined: 2 Oct 02
Posts: 404
Credit: 196,758
RAC: 0
United States
Message 124325 - Posted: 16 Jun 2005, 20:46:47 UTC


One of my XPs improved in FP speed, seven decreased slightly. All improved in Integer speed.

My A64 remained basically the same on both.

Thanks.

ID: 124325 · Report as offensive
Profile Sir Ulli
Volunteer tester
Avatar

Send message
Joined: 21 Oct 99
Posts: 2246
Credit: 6,136,250
RAC: 0
Germany
Message 124340 - Posted: 16 Jun 2005, 22:14:21 UTC

thanks to Tetsuji Maverick Rai

my P4 3.2 at 2.0 2 2 5 Twinmos Syster, and an I875 Mobo

http://setiathome.berkeley.edu/results.php?hostid=56765&offset=20

and my Athlon64 3.200+ on an S754 Mobo with the good Twinmos BH5 Chips also running at 2.0 2 2 5

http://setiathome.berkeley.edu/results.php?hostid=318394

that means

my P4 do 2 WUs at 1:45, and my Athlon64 about 1:12 :)

Greetings from Germany NRW
Ulli S@h Berkeley's Staff Friends Club m7 ©






ID: 124340 · Report as offensive
Divide Overflow
Volunteer tester
Avatar

Send message
Joined: 3 Apr 99
Posts: 365
Credit: 131,684
RAC: 0
United States
Message 124351 - Posted: 16 Jun 2005, 22:31:30 UTC - in response to Message 124325.  


One of my XPs improved in FP speed, seven decreased slightly. All improved in Integer speed.
My A64 remained basically the same on both.


This was a change to the SETI application, not the BOINC core client. There would have been no changes to your benchmark scores, just to your SETI@home processing times! Take a look and see if those systems crunch a WU any faster with the new YAOSCW r8.1 application running.

ID: 124351 · Report as offensive
j2satx
Volunteer tester

Send message
Joined: 2 Oct 02
Posts: 404
Credit: 196,758
RAC: 0
United States
Message 124358 - Posted: 16 Jun 2005, 23:06:44 UTC


I was not clear enough in my message.

The lower benchmarks were a result of implementing the "optimized" 4.45.

I am also running the "optimized" SETI client. I will have to review the processing times when several WUs have been processed.

Thanks.
ID: 124358 · Report as offensive
Profile Steve Cressman
Volunteer tester
Avatar

Send message
Joined: 6 Jun 02
Posts: 583
Credit: 65,644
RAC: 0
Canada
Message 124480 - Posted: 17 Jun 2005, 4:34:16 UTC - in response to Message 124358.  


I was not clear enough in my message.

The lower benchmarks were a result of implementing the "optimized" 4.45.

I am also running the "optimized" SETI client. I will have to review the processing times when several WUs have been processed.

Thanks.


Then why did you not post that in the thread about the optimized 4.45?
Thread: http://setiathome.berkeley.edu/forum_thread.php?id=15633

98SE XP2500+ @ 2.1 GHz Boinc v5.8.8

And God said"Let there be light."But then the program crashed because he was trying to access the 'light' property of a NULL universe pointer.
ID: 124480 · Report as offensive
Ned Slider

Send message
Joined: 12 Oct 01
Posts: 668
Credit: 4,375,315
RAC: 0
United Kingdom
Message 124519 - Posted: 17 Jun 2005, 9:51:50 UTC

The K version is good on my Duron (Morgan core) and a couple results have now validated. Slight increase in speed, I've not run a benchmark so I can't say exactly how much.

I will migrate the rest over this weekend :)

Good work Mav,

Ned

*** My Guide to Compiling Optimised BOINC and SETI Clients ***
*** Download Optimised BOINC and SETI Clients for Linux Here ***
ID: 124519 · Report as offensive
Profile -= Vyper =-
Volunteer tester
Avatar

Send message
Joined: 5 Sep 99
Posts: 1652
Credit: 1,065,191,981
RAC: 2,537
Sweden
Message 124545 - Posted: 17 Jun 2005, 12:22:24 UTC
Last modified: 17 Jun 2005, 12:23:31 UTC

Athlon64 3200+ @ 4100+ speeds.
DualChannel Memory @ 208 Mhz 2-2-2-6-1

73707066 17751801 14 Jun 2005 20:03:21 UTC 17 Jun 2005 11:16:58 UTC Over Success Done 3,230.72 15.31 35.00
73681344 17745500 14 Jun 2005 19:03:14 UTC 17 Jun 2005 11:15:56 UTC Over Success Done 3,756.05 17.80 27.97
73649898 17737936 14 Jun 2005 18:03:07 UTC 17 Jun 2005 10:20:21 UTC Over Success Done 3,230.91 15.31 pending

Two results returned with around 3200 secs and one with 3756 secs with the newest build of YAOSCW.

Thanks Tetsuji yet again!

_________________________________________________________________________
Addicted to SETI crunching!
Founder of GPU Users Group
ID: 124545 · Report as offensive
Profile Prognatus

Send message
Joined: 6 Jul 99
Posts: 1600
Credit: 391,546
RAC: 0
Norway
Message 124606 - Posted: 17 Jun 2005, 16:23:21 UTC - in response to Message 124181.  
Last modified: 17 Jun 2005, 16:25:47 UTC

YAOSCW-P-r8.1.zip - QxP = P4 SSE3
YAOSCW-N-r8.1.zip - QxN = P4 SSE2 only
YAOSCW-B-r8.1.zip - QxB = Pentium M SSE2
YAOSCW-W-r8.1.zip - QxW = P4 SSE2 AMD64
YAOSCW-K-r8.1.zip - QxK = P3 SSE AMD-XP
YAOSCW-0-r8.1.zip - No Option = P2 MMX

According to the horse's mouth, AMD supports SSE3 instructions on...:

Athlon 64
Athlon 64 X2
Opteron

And Athlon 64 FX has SSE2 support.

I didn't check any others, so I don't know if there are more.

ID: 124606 · Report as offensive
Heffed
Volunteer tester

Send message
Joined: 19 Mar 02
Posts: 1856
Credit: 40,736
RAC: 0
United States
Message 124633 - Posted: 17 Jun 2005, 17:38:34 UTC - in response to Message 124606.  

According to the horse's mouth, AMD supports SSE3 instructions on...:

Athlon 64

The SSE3 versions don't work on my Athlon64 FX-53.
ID: 124633 · Report as offensive
Profile Prognatus

Send message
Joined: 6 Jul 99
Posts: 1600
Credit: 391,546
RAC: 0
Norway
Message 124636 - Posted: 17 Jun 2005, 19:35:20 UTC - in response to Message 124633.  

No, the FX series only support SSE2 so far it seems. (see below)

ID: 124636 · Report as offensive
Heffed
Volunteer tester

Send message
Joined: 19 Mar 02
Posts: 1856
Credit: 40,736
RAC: 0
United States
Message 124637 - Posted: 17 Jun 2005, 19:38:15 UTC - in response to Message 124636.  

No, the FX series only support SSE2 so far it seems. (see below)

That's what happens when I post without having any sleep. ;) I totally missed that part about the FX's.
ID: 124637 · Report as offensive
Divide Overflow
Volunteer tester
Avatar

Send message
Joined: 3 Apr 99
Posts: 365
Credit: 131,684
RAC: 0
United States
Message 124645 - Posted: 17 Jun 2005, 20:20:18 UTC
Last modified: 17 Jun 2005, 20:28:15 UTC

SSE3 was added to the latest core versions for Athlon 64, Athlon 64 FX and Opteron. Like Intel, it's hard to know what your CPU really supports, since they've changed featuresets within CPU classifications. At tool like CPU-Z is great for checking the details.

Here's an FX-55 with SSE3 support, San Deigo core here.


ID: 124645 · Report as offensive
Ned Slider

Send message
Joined: 12 Oct 01
Posts: 668
Credit: 4,375,315
RAC: 0
United Kingdom
Message 124695 - Posted: 17 Jun 2005, 23:32:05 UTC

Just a heads up - for my amd64 linux client, adding SSE3 support for the latest Athlon64 Venice cores (that support it) made no difference over the SSE2 enabled version. I believe Metod and Maverick have seen the same thing on their Linux Intel P4 clients. So I wouldn't be surprised if the same were true for Windows clients. Don't expect a huge improvement (or any for that matter) from an SSE3 enabled client (over SSE2) just because your CPU supports it.

Ned

*** My Guide to Compiling Optimised BOINC and SETI Clients ***
*** Download Optimised BOINC and SETI Clients for Linux Here ***
ID: 124695 · Report as offensive
1 · 2 · 3 · 4 . . . 10 · Next

Message boards : Number crunching : YAOSCW r8.1 (optimized client for Windows)


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.