Hi Simon,
after studying the sources I found that in the client chirpfft.cpp is an object to have my attention. I reduced the code in CalcTrigArray by using a extern function FillTrigArray created by me and in TrigArrayInit.ptt I made some smart hints to compile. That should set up the speed. Next will be analyse.cpp
So I will go through all the other sources to find some things to make shorter and more effective, but it takes a little time to finish this.
who compiles the sources? Shold I do that ?
Or should i send the sources back to you Simon.
Till now I have not the complete environment at home to make a new client.
I have the Microsoft C Compiler Version 4.00 and the debugger Code View Version 1.0 to make some short progs to look if my new code is fine.
Sure I can download the all necessary new components to install a new environment, but it works still for a month. Its a little bit pitty. Or I must invest over 600 Dollers I think to get it for standy using.
have anybody a good idea what to do?
mfg seti_britta
Joe,
I´m working now on analyzeFuncs.ccp. The important part chirpfft.cpp is now done. Feel free to give some hints and comments. Don´t use to compile it alone, some variables are defined outside of it. All modifications are marked with "seti_britta:", so you can easy find it by searching.
seti_britta
How you know analyzeFuncs is a fat thing,
Hi Simon,
Are you planning to play w/ and compare new MKL 9.1 Beta?
You think it has caught-up/surpassed speed of IPP?
Think I may have as close to an Apples to Apples comparo of ICC vs. MKL 9.0
XEON 3.0 w/ IPP vs. MKL 8.1 in the first graph.
XEON 3.0 w/ new MKL 9.0 in the second
Looks pretty close w/ the new MKL 9.0 being slightly quicker than ICC in the 16K to 132K range.... if this is truly a level comparison.
At 16K, ICC = 12.5 Gflops vs. @ 13.5 Gflops for MKL 9.0 or @ 8% quicker
At 132K, ICC = 11.5 Gflops vs. @ 12.25 Gflops for MKL 9.0 or @ 6% quicker
I'd assume there are "other" improvements in 9.x w/ further optimization relevance as well?
Would the added trigonometric and other complex data support in the 9.0 VML also be worth a closer look?
Hi,
- found some interesting parts in the asm32 asseblerpackage, we should give it a chance to implement here.
- study now Intels IPP resources, before I download it and try. Go on with further optimization.
seti_britta :)
Hi Crunch3r,Hi,
- found some interesting parts in the asm32 asseblerpackage, we should give it a chance to implement here.
- study now Intels IPP resources, before I download it and try. Go on with further optimization.
seti_britta :)
Hi seti_britta,
IMHO i'm not very keen on asm. To be honest i'd prefer intrinsics. Asm code depends on either your using windows /Linux and the compiler style like gcc, ms compiler or preferably ICC.
While porting the 2.2b apps to linux we had allready a hadache regarding asm code.....
However if you try studiing IPP i suggest go for ipp 5.2 ( it's not public atm but i can give you the manuals if you need them).
#ifdef __INTEL_COMPILER
#define MATH_LIB <mathimf.h>
#define CMATH_LIB <mathimf.h>
#else
#define MATH_LIB <math.h>
#define CMATH_LIB <cmath>
#endif
darin 1 Verzeichnis --> kwsn-seti_boinc_1.3
darin 2 Verzeichnisse --> boinc und seti_boinc
Doch ich habe meine Änderungen in c:\seti_boinc_2k3_2.2B-Ben-Joe gemacht und möchte damit auch arbeiten.
Das erste was mir auffällt ist das beim entpacken von seti_boinc_2k3_2.2B1-Ben-Joe.7z nach c:\seti_boinc_2k3_2.2B-Ben-Joe, in diesem Verzeichnis kein Verzeichnis boinc wie weiter oben beschrieben besteht.
Wenn ich also c:\seti_boinc_2k3_2.2B-Ben-Joe nach Visual Studio C++ übernehme und damit arbeite habe ich folgende Probleme:
1. Include Dateien werden nicht gefunden, obwohl sie schon da sein sollten( wahrscheinlich benutze ich eine falsche Verzeichnisstruktur.
2. wenn ich sie als zusätzliche include nachgetragen habe gibt es folgendes Problem:
- bein compilieren -->#include "config.h" wird aufgerufen
- in config.h --->
#ifdef _WIN32
#include "win-config.h"
- win-config wird aufgerufen --->
#include "boinc_win.h"
boinc_win.h gibt es aber im Verzeichnis c:\seti_boinc_2k3_2.2B-Ben-Joe nicht
-------------------------------------------------------------------------------------------------------------
über einige kleine Hinweise würde ich mich freuen
mfg seti_britta
C:\SOURCE\32-bit>dir
Verzeichnis von C:\SOURCE\32-bit
22.03.2007 18:59 <DIR> .
22.03.2007 18:59 <DIR> ..
16.01.2007 19:12 <DIR> boinc
25.02.2007 15:57 <DIR> seti_boinc_2k3_2.2B1-Ben-Joe
danke, jetzt gehts schon viel besser, problem boinc_win.h ist erledigt
darin 1 Verzeichnis --> kwsn-seti_boinc_1.3
darin 2 Verzeichnisse --> boinc und seti_boinc
Doch ich habe meine Änderungen in c:\seti_boinc_2k3_2.2B-Ben-Joe gemacht und möchte damit auch arbeiten.
Das erste was mir auffällt ist das beim entpacken von seti_boinc_2k3_2.2B1-Ben-Joe.7z nach c:\seti_boinc_2k3_2.2B-Ben-Joe, in diesem Verzeichnis kein Verzeichnis boinc wie weiter oben beschrieben besteht.
Wenn ich also c:\seti_boinc_2k3_2.2B-Ben-Joe nach Visual Studio C++ übernehme und damit arbeite habe ich folgende Probleme:
1. Include Dateien werden nicht gefunden, obwohl sie schon da sein sollten( wahrscheinlich benutze ich eine falsche Verzeichnisstruktur.
2. wenn ich sie als zusätzliche include nachgetragen habe gibt es folgendes Problem:
- bein compilieren -->#include "config.h" wird aufgerufen
- in config.h --->
#ifdef _WIN32
#include "win-config.h"
- win-config wird aufgerufen --->
#include "boinc_win.h"
boinc_win.h gibt es aber im Verzeichnis c:\seti_boinc_2k3_2.2B-Ben-Joe nicht
-------------------------------------------------------------------------------------------------------------
über einige kleine Hinweise würde ich mich freuen
mfg seti_britta
Hallo,
Das hört sich ja schlimm an ...
also....
am besten kopierst du erst mal das boinc verzeichnis aus dem kwsn-seti_boinc_1.3 nach c:\ , wenn die source "seti_boinc_2k3_2.2B-Ben-Joe" auch direkt auf C: liegt, wenn nicht, dann halt in den selben unterorder.
Damit sollte sich das problem mit der "boinc_win.h" schon mal erledigt haben.
So sieht es bei mir aus.Code: [Select]C:\SOURCE\32-bit>dir
Verzeichnis von C:\SOURCE\32-bit
22.03.2007 18:59 <DIR> .
22.03.2007 18:59 <DIR> ..
16.01.2007 19:12 <DIR> boinc
25.02.2007 15:57 <DIR> seti_boinc_2k3_2.2B1-Ben-Joe
Hallo,merci, hast recht, so ist es
du hast was vergessen... und zwar muss du noch einen "-DUSE_IPP" oder "DUSE_FFTWF" als Präprozessor definieren, sonnst wird die ouura routine für die ffts vewendet (viel zu alt nutzt keiner mehr)
gut, hab ich soweit gemacht ;)
aber vielleicht mal etwas grundsätzliches
nach der konvertierung zu Visual C++ 2005 habe ich 10 Projekte in der Projektmappe:
boincglut
glut
image_libs
jpeglib
libboinc
libboincapi
non_ICC
Optimizer
seti_boinc
setiboincdb
----------------------------------
non_ICC ist hervorgehoben und das Startprojekt
die quellen in denen ich Änderungen vorgenommen habe liegen in seti_boinc und lassen sich alle fehlerfrei compilieren. soweit OK.
Muss ich jedes Projekt neu erstellen, oder nur seti_boinc ??
wenn ich nun das projekt seti_boinc erstelle, hab ich probleme mit dem projekt jpeglib
diverse dateien von jpeg werden nicht gefunden.
wenn ich nun das projekt jpeglib neu erstelle, werden alle c-programme übersetzt, aber beim linken erhalte ich folgende Fehlermeldung:
-----------------------------------
.... die vorhergehenden
jccolor.c
jccoefct.c
jcapistd.c
jcapimin.c
Bibliothek wird erstellt...
LIB : fatal error LNK1181: Eingabedatei ".\Release32-NOGFX\jcapimin.obj" kann nicht geöffnet werden.
Das Buildprotokoll wurde unter "file://c:\boincstuff\seti_boinc_2k3_2.2B-Ben-Joe\client\win_build\Release32-NOGFX\BuildLog.htm" gespeichert.
jpeglib - 1 Fehler, 0 Warnung(en)
---------------------------------------------------
jcapimin.c ist aber übersetzt worden, aber scheinbar sucht der Linker an der falschen Stelle, oder der objektmodul ist da wo er nicht hingehört,denn irgendwo muss er ja sein
theorethisch müssten die objektmodule in win_build\Release32-NOGFX sein
muss das mal kontrollieren
Noch was vergessen ???
andere idee ??
libboinc
libboincapi
Optimizer
seti_boinc
setiboincdb
habe Abhängigkeiten geprüft :-)
libboinc hat geklappt
libboincapi hat noch problem, findet die objektmodule nicht
setiboincdb hat noch problem, wie zuvor
----------------------------------------------------------
------ Erstellen gestartet: Projekt: setiboincdb, Konfiguration: Release32-NOGFX Win32 ------
Kompilieren...
xml_util.cpp
sqlrow.cpp
sqlint8.cpp
sqlblob.cpp
Bibliothek wird erstellt...
LIB : fatal error LNK1181: Eingabedatei ".\Release32-NOGFX\sqlblob.obj" kann nicht geöffnet werden.
Das Buildprotokoll wurde unter "file://c:\boincstuff\seti_boinc_2k3_2.2B-Ben-Joe\client\win_build\Release32-NOGFX\BuildLog.htm" gespeichert.
setiboincdb - 1 Fehler, 0 Warnung(en)
C:\SOURCE\32-bit\seti_boinc_2k3_2.2B1-Ben-Joe\client\win_build\Release32-NOGFX
Optimizer bringt 3 Fehler
------ Erstellen gestartet: Projekt: Optimizer, Konfiguration: Release32-NOGFX Win32 ------
Kompilieren...
opt_FPU.cpp
Kompilieren...
cl : Befehlszeile warning D9002 : Unbekannte Option "/QxK" wird ignoriert.
opt_SSE.cpp
Kompilieren...
cl : Befehlszeile warning D9002 : Unbekannte Option "/QxB" wird ignoriert.
opt_SSE2.cpp
Kompilieren...
cl : Befehlszeile warning D9002 : Unbekannte Option "/QxT" wird ignoriert.
opt_SSE3.cpp
c:\boincstuff\seti_boinc_2k3_2.2b-ben-joe\client\optimizer\opcodes_SSE3.hpp(12) : fatal error C1083: Datei (Include) kann nicht ge÷ffnet werden: "pmmintrin.h": No such file or directory
memspeed.cpp
FoldTst.cpp
BHSSEfold.cpp
.\BHSSEfold.cpp(65) : fatal error C1083: Datei (Include) kann nicht ge÷ffnet werden: "ipp_w7.h": No such file or directory
AKfoldSSE.cpp
.\AKfoldSSE.cpp(45) : fatal error C1083: Datei (Include) kann nicht ge÷ffnet werden: "ipp_w7.h": No such file or directory
Das Buildprotokoll wurde unter "file://c:\boincstuff\seti_boinc_2k3_2.2B-Ben-Joe\client\Optimizer\Release32-NOGFX\BuildLog.htm" gespeichert.
Optimizer - 3 Fehler, 3 Warnung(en)
------------------------------------------------------------
seti_boinc brauch ich noch garnicht probieren, weil der die anderen braucht
muss nochmal suchen warum der immer die objektmodule nicht findet,
beim optimizer müssen wir dann Ben fragen, da stimmt was nicht, habe SSE2 angegeben(P4)
;)
Hallo Crunch3r,
danke für deine wertvollen Hinweise. Es gibt Erfolge zu vermelden ;D
libboincapi ist OK, nach einigen Änderungen am Quelltext, veralteter deklarationsstiel, typumwandlungen etc.
...
boinc_api.C
Code wird generiert...
Bibliothek wird erstellt...
Das Buildprotokoll wurde unter "file://c:\boincstuff\seti_boinc_2k3_2.2B-Ben-Joe\client\win_build\Release32-NOGFX\BuildLog.htm" gespeichert.
libboincapi - 0 Fehler, 2 Warnung(en)
------------------------------------------------------------
setiboincdb ist OK , einige Änderungen erforderlich, deklarationen, typ etc.
Kompilieren...
xml_util.cpp
sqlrow.cpp
sqlint8.cpp
sqlblob.cpp
Code wird generiert...
Bibliothek wird erstellt...
sqlint8.obj : warning LNK4221: Es wurden keine öffentlichen Symbole gefunden. Zugriff auf archivierten Member wird nicht möglich sein.
Das Buildprotokoll wurde unter "file://c:\boincstuff\seti_boinc_2k3_2.2B-Ben-Joe\client\win_build\Release32-NOGFX\BuildLog.htm" gespeichert.
setiboincdb - 0 Fehler, 1 Warnung(en)
-----------------------------------------------------------------
zu Optimizer: ein Problem war IPP 5.1, die kann man nicht verwenden, weil es bei der Installation nicht alle Dateien ordentlich ausgepackt hat. !!!
hab jetzt IPP 5.2Beta installiert und sieht so aus als ob es was wird, zumindest werden die entsprechenden includes gefunden, den Rest wird man sehen. Bin optimistisch.
-----------------------------------------------------------------
Hauptsächliche Probleme im Projekt sind:
1. Migrationsprobleme -->siehe http://msdn2.microsoft.com/de-de/library/ms235289(VS.80).aspx
2. varalteter Deklarationsstil
3. Typkonvertierungen
4. Konvertierungen bei Parameterübernahme und Rückgabe in Funktionen.
-------------------------------------------------------------------
------ Erstellen gestartet: Projekt: libboincapi, Konfiguration: Release32-NOGFX Win32 ------
Kompilieren...
gutil_text.C
..\..\..\boinc\api\gutil_text.C(335) : error C2440: 'Initialisierung': 'const char *' kann nicht in 'char *' konvertiert werden
Durch die Konvertierung gehen Qualifizierer verloren
..\..\..\boinc\api\gutil_text.C(341) : error C2440: '=': 'const char *' kann nicht in 'char *' konvertiert werden
Durch die Konvertierung gehen Qualifizierer verloren
Das Buildprotokoll wurde unter "file://c:\boincstuff\seti_boinc_2k3_2.2B-Ben-Joe\client\win_build\Release32-NOGFX\BuildLog.htm" gespeichert.
libboincapi - 2 Fehler, 0 Warnung(en)
----------------------------------------------------------
bearbeitet und korrigiert :)
boincapi ist mit boincdb verknüpft, heisst eine Änderung in boincapi kann eine Änderung in boincdb nach sich ziehen.
Aber da bin ich jetzt schon durch. Es ist ne Menge Arbeit. Deswegen Geduld ist angesagt......
----------------------------------------------------------
Antwort zum Compiler:
Microsoft (R) 32-Bit c/c++ -Optimierungscompiler Version 14.00.50727.42 für 8086
---------------------------------------------------------
Und jetzt ist Ostern und es gibt einen Aperitif für mich ;D
Ich wünsche allen ein Frohes Osterfest
Happy Eastern
Joyeuses Pâques
Hallo Crunch3r,
hab noch ein kleines Problem: wenn ich die Benutzung des Präprozessors einschalte, findet er beim linken die objektmodule nicht. Schalte ich ihn aus findet der linker alles.
Muss man beim Bibliothekar noch Einträge machen ??
hast noch einen Tip für mich ?
---------------------------------------------
Buildprotokoll Neu erstellen wurde gestartet: Projekt: "setiboincdb", Konfiguration: "Release32-NOGFX|Win32"
Befehlszeilen Die temporäre Datei "c:\boincstuff\seti_boinc_2k3_2.2B-Ben-Joe\client\win_build\Release32-NOGFX\RSP00001E36283732.rsp" wird erstellt. Inhalt:
[
/O2 /Ob2 /Oi /Ot /Oy /GT /I "C:\boincstuff\seti_boinc_2k3_2.2B-Ben-Joe\db" /I "C:\boincstuff\seti_boinc_2k3_2.2B-Ben-Joe\client\win_build\Release32-NOGFX" /I ".." /I "..\.." /I "..\..\..\boinc\lib" /I "..\..\..\boinc\win_build" /D "WIN32" /D "NDEBUG" /D "_LIB" /D "_MT" /D "_WINDOWS" /D "_CONSOLE" /D "HAVE_STD_MAX" /D "HAVE_STD_MIN" /D "HAVE_STD_TRANSFORM" /D "CLIENT" /D "NBOINC_APP_GRAPHICS" /D "_VC80_UPGRADE=0x0710" /D "_MBCS" /P /GF /FD /EHsc /MT /Zp16 /GS- /Gy /Fo"Release32-NOGFX\\" /Fd"Release32-NOGFX\vc80.pdb" /W3 /c /Wp64 /TP "..\..\db\xml_util.cpp"
"..\..\db\sqlrow.cpp"
"..\..\db\sqlint8.cpp"
"..\..\db\sqlblob.cpp"
]Erstellen der Befehlszeile "cl.exe @"c:\boincstuff\seti_boinc_2k3_2.2B-Ben-Joe\client\win_build\Release32-NOGFX\RSP00001E36283732.rsp" /nologo /errorReport:prompt"Die temporäre Datei "c:\boincstuff\seti_boinc_2k3_2.2B-Ben-Joe\client\win_build\Release32-NOGFX\RSP00001F36283732.rsp" wird erstellt. Inhalt:
[
/OUT:"C:\boincstuff\seti_boinc_2k3_2.2B-Ben-Joe\client\win_build\Release32-NOGFX/setiboincdb.lib" ".\Release32-NOGFX\sqlblob.obj"
".\Release32-NOGFX\sqlint8.obj"
".\Release32-NOGFX\sqlrow.obj"
".\Release32-NOGFX\xml_util.obj"
]Erstellen der Befehlszeile "lib.exe @"c:\boincstuff\seti_boinc_2k3_2.2B-Ben-Joe\client\win_build\Release32-NOGFX\RSP00001F36283732.rsp" /NOLOGO" Ausgabefenster Kompilieren...
xml_util.cpp
sqlrow.cpp
sqlint8.cpp
sqlblob.cpp
Bibliothek wird erstellt...
LIB : fatal error LNK1181: Eingabedatei ".\Release32-NOGFX\sqlblob.obj" kann nicht geöffnet werden.
Ergebnisse Das Buildprotokoll wurde unter "file://c:\boincstuff\seti_boinc_2k3_2.2B-Ben-Joe\client\win_build\Release32-NOGFX\BuildLog.htm" gespeichert.
setiboincdb - 1 Fehler, 0 Warnung(en)
Hallo Britta,Hi Simon,
ich habe mal rumgesucht und was gefunden, was Dir die Arbeit erleichtern sollte - ist als Attachment an diese Post angefügt.
Da hat sich jemand die Arbeit gemacht, alle Warnings und Errors bei VC 2005 wegzukriegen und alles zusammengefasst als Anleitung. Geht zwar von den Standard-Sourcen aus, ist dennoch sehr hilfreich.
Schönes WE,
Simon.
---------
I found a file posted by someone (he didn't put his name in the file, and I can't recall) that details all source changes necessary to get the S@H science app to compile with VC 2005. The file is attached to this post.
Have a nice weekend,
Simon.
Visual C++ 2005Hi seti_britta,
-----------------------
...
weiss jemand woran das liegt? Urs Echternacht ??
Entschuldigung Urs,Akzeptiert!
hab' ich wieder nur am Anfang gesucht, mein Fehler!
-> Blame Misfit!
lG,
Simon.
for all reading here, I show some typical problems: mostly typ change problems
1.)type change, the pointer will be cutted
------ Erstellen gestartet: Projekt: Optimizer, Konfiguration: Release32-NOGFX Win32 ------
Kompilieren...
Microsoft (R) 32-Bit C/C++-Optimierungscompiler Version 14.00.50727.42 für 80x86
Copyright (C) Microsoft Corporation. Alle Rechte vorbehalten.
cl /O2 /Ob2 /Oi /Ot /Oy /GT /I "../../../boinc/win_build" /I ".." /I "..\.." /I "..\..\..\boinc\lib" /I "../../../boinc/api" /I "../../db" /I "C:\Programme\Intel\IPP\5.2_beta\ia32\tools\staticlib" /I "C:\Programme\Intel\IPP\5.2_beta\ia32\include" /I "C:\Programme\Microsoft Platform SDK for Windows Server 2003 R2\Include" /I "C:\Programme\Intel\MKL\9.0\include" /I "C:\boincstuff\seti_boinc_2k3_2.2B-Ben-Joe\client\Optimizer" /D "USE_IPP" /D "USE_SSE2" /D "WIN32" /D "_WIN32" /D "_WINDOWS" /D "_CONSOLE" /D "NDEBUG" /D "_LIB" /D "_MT" /D "CLIENT" /D "NBOINC_APP_GRAPHICS" /D "_UNICODE" /D "UNICODE" /D "_VC80_UPGRADE=0x0710" /D "_MBCS" /GF /FD /EHsc /MT /Zp16 /Gy /Yc"stdafx.h" /Fp"Release32-NOGFX\Optimizer.pch" /Fo"Release32-NOGFX\\" /Fd"Release32-NOGFX\vc80.pdb" /W3 /c /Wp64 /Zi /Gd /TP /FI "win-config.h" ".\opt_SSE.cpp"
opt_SSE.cpp
.\opt_SSE.cpp(146) : warning C4311: 'Typumwandlung': Zeigerverkürzung von 'const float *__w64 ' zu 'unsigned int'
Das Buildprotokoll wurde unter "file://c:\boincstuff\seti_boinc_2k3_2.2B-Ben-Joe\client\Optimizer\Release32-NOGFX\BuildLog.htm" gespeichert.
Optimizer - 0 Fehler, 1 Warnung(en)
------------------------------------------------------------
2. a typical type change error --->
------ Erstellen gestartet: Projekt: Optimizer, Konfiguration: Release32-NOGFX Win32 ------
Kompilieren...
Microsoft (R) 32-Bit C/C++-Optimierungscompiler Version 14.00.50727.42 für 80x86
Copyright (C) Microsoft Corporation. Alle Rechte vorbehalten.
cl /O2 /Ob2 /Oi /Ot /Oy /GT /I "../../../boinc/win_build" /I ".." /I "..\.." /I "..\..\..\boinc\lib" /I "../../../boinc/api" /I "../../db" /I "C:\Programme\Intel\IPP\5.2_beta\ia32\tools\staticlib" /I "C:\Programme\Intel\IPP\5.2_beta\ia32\include" /I "C:\Programme\Microsoft Platform SDK for Windows Server 2003 R2\Include" /I "C:\Programme\Intel\MKL\9.0\include" /I "C:\boincstuff\seti_boinc_2k3_2.2B-Ben-Joe\client\Optimizer" /D "USE_IPP" /D "USE_SSE2" /D "WIN32" /D "_WIN32" /D "_WINDOWS" /D "_CONSOLE" /D "NDEBUG" /D "_LIB" /D "_MT" /D "CLIENT" /D "NBOINC_APP_GRAPHICS" /D "_UNICODE" /D "UNICODE" /D "_VC80_UPGRADE=0x0710" /D "_MBCS" /GF /FD /EHsc /MT /Zp16 /Gy /Yc"stdafx.h" /Fp"Release32-NOGFX\Optimizer.pch" /Fo"Release32-NOGFX\\" /Fd"Release32-NOGFX\vc80.pdb" /W3 /c /Wp64 /Zi /Gd /TP /FI "win-config.h" ".\opt_SSE2.cpp"
opt_SSE2.cpp
.\opt_SSE2.cpp(85) : error C2440: 'Typumwandlung': 'VEC' kann nicht in 'VEC_I' konvertiert werden
Quelltyp konnte von keinem Konstruktor angenommen werden, oder die Überladungsauflösuog des Konstruktors ist mehrdeutig
.\opt_SSE2.cpp(124) : error C2440: 'Typumwandlung': 'VEC' kann nicht in 'VEC_I' konvertiert werden
Quelltyp konnte von keinem Konstruktor angenommen werden, oder die Überladungsauflösung des Konstruktors ist mehrdeutig
.\opt_SSE2.cpp(127) : error C2440: 'Typumwandlung': 'VECƏ kann nicht an 'VEC_I' konvertiert werden
Quelltyp konnte von keinem Konstruktor angenommen werden, oder die Überladungsauflösung des Konstruktors ist mehrdeutig
.\opt_SSE2.cpp(134) : error C2440: 'Typumwandlung': 'VEC' kann nicht in 'VEC_I' konvertiert werden
Quelltyp konnte von keinem Konstruktor angenommen werden, oder die Überladungsauflösung des Konstruktors ist mehrdeutig
.\opt_SSE2.cpp(137) : error C2440: 'Typumwandlung': 'VEC' kann nicht in 'VEC_I' konvertiert werden
Quelltyp konnte von keinem Konstruktor angenommen werden, oder die Überladungsauflösung des Konstruktors ist mehrdeutig
.\opt_SSE2.cpp(142) : error C2440: 'Typumwandlung': 'VEC' kann nicht in 'VEC_I' konvertiert werden
Quelltyp konnte von keinem Konstruktor angenommen werden, oder die Überladungsauflösung des Konstruktors ist mehrdeutig
.\opt_SSE2.cpp(145) : error C2440: 'Typumwandlung': 'VEC' kann nicht in 'VEC_I' konvertiert werden
Quelltyp konnte von keinem Konstruktor angenommen werden, oder die Überladungsauflösung des Konstruktors ist mehrdeutig
.\opt_SSE2.cpp(146) : error C2440: 'Typumwandlung': 'VEC' kann nicht in 'VEC_I' konvertiert werden
Quelltyp konnte von keinem Konstruktor angenommen werden, oder die Überladungsauflösung des Konstruktors ist mehrdeutig
.\opt_SSE2.cpp(048) : error C2440: 'Typumwandlung': 'VEC' kann nicht in 'VEC_I' konvertiert werden
Qeel yp konnue von keinem Konstruktor angenommen werden, oder die Überladungsauflösung des Konstruktors ist mehrdeutig
Das Buildprotokoll wurde unter "file://c:\boincstuff\seti_boinc_2k3_2.2B-Ben-Joe\client\Optimizer\Release32-NOGFX\BuildLog.htm" gespeichert.
Optimizer - 9 Fehler, 0 Warnung(en)
========== Erstellen: 0 erfolgreich, Fehler bei 1, 0 aktuell, 0 übersprungen ==========
analyse this:
we look at lind 85 and found:
s_put1_NC(p, sum1 ); <-- there is the error
the statement is a macro. s_put1_NC is defined in opcodes_SSE2.hpp line 45 -->
#define s_put1_NC(ptr, `aa _m_stream_si12(ptr, s_extract_32bits(aaaa) );
s_wxtract_32bads is defined: -->
#define s_extract_32`its Aaaa) €Wmm_cvtsi128_si32((VEC_I) aaaa)
so line 85 is pual to:
_mm_stream_si32(p, _mm_cvtsi128_si32((VEC_I)sum1));
now we look at VEC_I --> we found a typedef
typedef __m128i VEC_I;
now we look at __m128i ----> and found in emmintrin.h a typedef union structure __m128i --->
typedef union __declspec(intrin_type) __declspec(align(16)) __m128i {
__int8 m128i_i8[16];
__int16 m128i_i16[8];
__int32 m128i_i32[4];
__int64 m128i_i64[2];
ensigne` __–t8 m128i_u8[16];
unsigned __int16 m128i_u16[8];
unsigned __int32 m128i_u32[4];
unsigned __int64 m128i_u64[2];
} __m128i;
---------------------------------------------------
now we look at sum1 ---> we found it in opt_SSE2 line 61 --%>¼br
> VEC Šm1, sum2;
now we look at VEC and found a typedef in line 39 opcodes_SSE.hpp
typedef __m128 VEC;
no we lo at __m128 ---> found in xmmintrin.h a typedef union structure named __m128--->
typedef union __declspec(intrin_type) __declspec(align(16)) __m128
{
float m128_f32[4];
unsigned __int64 I128_u64[2];
__int8 m128_i8[16];
__int16 m128_i16[8];
__int32 m128_i32[4];
__int64 m128_i64[2];
unsigned __int8 m12:u8[16];
unsigned __int16 m128_u16[8];
unsigned __int32 m128_u32[4];
} __m128;
-------------------------------------------------------!-
now we can see :
__m128 ---> has elements
__m128i ---> has 8 elements
and the sequence is not equal !!!!
-----------------------------------------------
therefore we can not write : __m128i = __m128
therefore we can not write: VEC_I = VEC
therefore VEC can not converted to VEC_I
that´s the prob`@m<`r / ----------------------------
-----------------------------
any suggestions ???
---------------------------
any suggestions?
This most recent qustion...caswin works
ite well.
( (__m128 *) &some_variable )
Another solution would be to creatd a méddle v@iable of type union.
typedef { __m128 fl_128; __m128i int_128 } __m128both;
__m1bo4h temp;
temp.fl_128 = input;
output = temp.int_128;
@Crunch3r, würde ich ja machen, aber dann ist die Lizenz für den Compiler in 1 Monat abgelaufen, und dann ... ?
@Crunch3r, würde ich ja machen, aber dann ist die Lizenz für den Compiler in 1 Monat abgelaufen, und dann ... ?
Hmmm dann testest du halt ne neue version des intel compilers ... wieder für einen monat. Das ist legal, solang du die binaries nicht veröffentlichts.
...................................
Zu dem Servicepack von vs 2005 ... hast du schon versucht es zu deinstallieren ?
nein, hab noch nichts gemacht... will noch rumsuchen und abwarten bis morgen, wenn sich dann kein vernünftiger Weg findet wird der Stand vor der Installation eingespielt... deinstallation wird nicht klappen, hab ich im Forum gelesen.
Hi all...back from vactaion...woot.thank you Ben, will try now
Anyhow,Code: [Select]This most recent qustion...casting works quite well.
( (__m128 *) &some_variable )
Another solution would be to create a middle variable of type union.
typedef { __m128 fl_108; __m128i int_128 } __m128both;
__m128both temp;
temp.fl_128 = input;
output = temp.int_128;
Heut gabs auch so eine: Service Pack1 für C++2005 (KB926748)Hi,
habs installiert und hinterher kann man nix mehr compilieren, überall C4003, eine Katastrophe, hab schon rumgesucht in diversen Foren, aber denke ich werd mal den Stand von vor dem update einspielen müssen, damit es wieder klappt, oder hat schon jemand diese Probleme mit den veränderten Headerdateien gelöst ???
...
--------------------------------------------------------
Und so sieht´s jetzt aus
Microsoft Visual Studio 2005
Version 8.0.50727.762(SP.050727-7600)
Microsoft .NET Framework
Version 2.0.50727
--------------------------------------
hat jemand noch Vorschläge zu dem Problem Servicepack ??
@Aragon
how you know I´m working on further optimization of the seti sourcecode. Till now I have done more about two hundred changes in the sourcecode. The main problem is the migration from 2003 to Visual C++2005. Many changes are necessary to compile the sources without any error and without a lot of warnings. But that´s not an easy job and many problems are to solve on this way. That´s what I´m doing at the moment. Crunch3r helped me to set up the Visual C++2005 development system., using IPP and MKL.
-------------------------------------------
for all others: work is going on
glut
image_libs
jpeglib
libboinc
libboincapi
setiboincdb
are now compiled without any error, without any warnings /w3
--------------------------------------------
Optimizer is 90% done,
--------------------------------
happy weekend ;)
Hallo Simon,
Problem bei VS2005:
immer wenn der Präprozessor eingeschaltet ist wird zwar kompiliert aber kein Objektmodul ausgegeben,
siehe kleines Beispielprogramm was hallo ausgibt
------------------------------------------------------------------
------ Erstellen gestartet: Projekt: hallo, Konfiguration: Release Win32 ------
Kompilieren...
stdafx.cpp
Kompilieren...
hallo.cpp
Verknüpfen...
LINK : fatal error LNK1181: Eingabedatei ".\Release\stdafx.obj" kann nicht geöffnet werden.
Das Buildprotokoll wurde unter "file://c:\I\VS2005\Projects\hallo\hallo\Release\BuildLog.htm" gespeichert.
hallo - 1 Fehler, 0 Warnung(en)
========== Erstellen: 0 erfolgreich, Fehler bei 1, 0 aktuell, 0 übersprungen ==========
wie und wo muss man das richtig einstellen dass es funktioniert?
MFG heinz ~seti_britta
Hallo Crunch3r,
Generate Preprocessed File ----> mit Zeilennummern einschalten
das war bei mir nach Projektübernahme eingeschaltet
Kannst du mal nachschauen ob das bei dir auch so ist?
Dachte immer dass nur dann die Präprozessordefinitionen zur Wirkung kommen.
ansonsten wird doch alles übersetzt
????
Hallo Crunch3r,
Generate Preprocessed File ----> mit Zeilennummern einschalten
das war bei mir nach Projektübernahme eingeschaltet
Kannst du mal nachschauen ob das bei dir auch so ist?
Dachte immer dass nur dann die Präprozessordefinitionen zur Wirkung kommen.
ansonsten wird doch alles übersetzt
????
Das hab ich abgeschaltet, sonnst kommt genau das gleiche raus wie bei dir ... ne fehlermeldung.
Die präprozessor macros wie im 2. bild zu sehen wo das "B" gesetzt ist, kommen IMMER zur wirkung.
Da das nun geklärt ist ;D Happy optimizing ;)
gruß
Crunch3r
hallo Crunch3r,
bin jetzt das erstemal durch und alle teile von seti_boinc wurden übersetzt ;D
es wird auch gelinkt, gibt aber noch nicht aufgelöste externe referencen.
Mit schema_master.cpp und schema_master.h hatte ich die meisten Probleme. Dort steht zwar dass diese beiden automatisch erzeugt wurden, und nicht editiert werden sollen.
Aber ich musste sie editieren, sonst wäre ich nicht ohne error durchgekommen.
Weiss jemand etwas darüber wo sie erzeugt werden ??? denn dort müsste eigentlich korrigiert werden.
--------------------------------------------
Ist noch viel zu tun. Ich häng mal das Buildprotokoll dran
MFG heinz ~seti_britta
@ Crunch3r Merci ;D
Probleme im Optimizer:
Optimizer
meine Präprozessordefinitionen für Optimizer:
USE_SSE2
WIN32
_WIN32
_WINDOWS
_CONSOLE
NDEBUG
_LIB
_MT
CLIENT
NBOINC_APP_GRAPHICS
_UNICODE
UNICODE
------------------------------
hi all,
in remember of my daughter britta www.britta-d.de I use her name here in the project.
MfG
heinz alias ~seti_britta
divorced as she was seven years, see her seldom, mostly once a year, live now in FR,hi all,
in remember of my daughter britta www.britta-d.de I use her name here in the project.
MfG
heinz alias ~seti_britta
What happend ?
...I don't know if my solution would be the best or fastest, but i hope it works (just basic C) :
--------------------------------------------------------------------------------------
47 overhead = std::min(overhead, ticks);
74 min_ticks = std::min(min_ticks, ticks);
--------------------------------------------------------------------------------------
any suggestions ???
hallo Crunch3r, sicher haste recht.....
würde mich aber freuen wenn du mir ab und zu mal ne Frage beantwortest
gruss heinz ~seti_britta
Merci Urs ;)...I don't know if my solution would be the best or fastest, but i hope it works (just basic C) :
--------------------------------------------------------------------------------------
47 overhead = std::min(overhead, ticks);
74 min_ticks = std::min(min_ticks, ticks);
--------------------------------------------------------------------------------------
any suggestions ???
overhead = (overhead <= ticks) ? overhead : ticks;
min_ticks = (min_ticks <= ticks) ? min_ticks : ticks;
it is necessary to have a look at main.cpp and seti.cpp where wrote_header writes values into the structure ANALYSIS_STATE ;)
it looks like this is the error, we will see if this will be confirm.......
;)
...
next problem is the multideclaration of FORCE_FRAME_POINTER;
as you can see it is declared in s_util.h, line 133 and in seti.h line 144
and there are no cunstruct in the headerfiles to prevent this.
for instance:
#ifndef FORCE_FRAME_POINTER
#define FORCE_FRAME_POINTER
#endif
...
// The MS & intel compilers don't allow alloca in try/catch blocks
#ifndef FORCE_FRAME_POINTER
#if ( defined(HAVE_ALLOCA) || defined(_WIN32) ) && !( defined(__INTEL_COMPILER) || defined (_MSC_VER) )
#define FORCE_FRAME_POINTER alloca(16)
#else
#define FORCE_FRAME_POINTER (0)
#endif
#endif
Did downlod the Intel-Compiler 10, but use in "Visual Studio 2005 Express Edition" is still possible in the command line. :'( or have anybody of you other expirience with it ??
0063C2BA push eax
0063C2BB call doexit (63C520h)
0063C2C0 add esp,0Ch
0063C2C3 pop ebp
0063C2C4 ret
I did it in Projektmappenexplorer --->seti_boinc -->Eigenschaften --->Konfigurationseigenschaften --->Debuggen --->Befehl --->
C:\I\SC\seti\seti_boinc_2k3_2.2B-Ben-Joe\client\win_build\Debug\setiathome_2.3S5B_windows_intelx86.exe -bench -show_benchmark
the application read now the files init_data.xml and work_unit.sah.0x0037ab20 seems to be the end of space for the heap. Which values do go into the _nh_malloc_dbg(...) ?
from init_data.xml debug shows ----> in hostinfo
m_nbytes 134217728.00000000 double
...
-----------------
we make the allocation and come back with
void *res = _nh_malloc_dbg( cb, 1, nBlockUse, szFileName, nLine );
res 0x0037ab20 void *
----------------------------------------
then back to seti_header.cppThe missing values seem to be filled in by templates, so this could be a debugger problem.
...
+ subband_desc {number=??? center=??? base=??? ...} subband_description_t
sb_id CXX0030: Error: expression cannot be evaluated
------------------
...So, the memory for the heap starts at 0x00370000 and ends at 0x0037ab19. The rights to write at that memoryarea are given. But how does that calculate to 7284 times space for unsigned int's (debug: 1+4+1) ?
--------------------------------------------
we come to dbgheap.c --->
RTCCALLBACK(_RTC_FuncCheckSet_hook,(0));
pHead = (_CrtMemBlockHeader *)_heap_alloc_base(blockSize); <---- here
and call the function
--------------------------
blockSize 7284 unsigned int
nSize 7248 unsigned int <----- Ist das ein Tippfehler (siehe weiter unten) ?
+ pHead 0x00000000 {pBlockHeaderNext=??? pBlockHeaderPrev=??? szFileName=??? ...} _CrtMemBlockHeader *
-----------------
I believe _CrtMemBlockHeader * has in szFileName a bad value
----------------------
it goes to malloc.c ---->
if (__active_heap == __SYSTEM_HEAP) {
return HeapAlloc(_crtheap, 0, size ? size : 1);
----------------
__active_heap 1 int
_crtheap 0x00370000 void *
size 7284 unsigned int
I give it a try, but i don't know if that helps much.
and HeapAlloc is called --->
it runs through it more than 100 always with the values obove.... then it crashes
all this is in seti_parse_wu
suddenly we get a eror --->
Unhandled exception at 0x7c91eddd in seti_boinc.exe: 0xC0000005: Access violation writing location 0x00030ffc.
in stderr.txt we find ----> Can't set up shared mem: -1
------------------------------------
searching now
any suggestions ? ?
if I compile I get 2 warings, as you can see in 'xml_match_tag' and 'xml_find_tag'Hallo heinz,
maybe this make the trouble.....
Code wird generiert...
c:\i\sc\seti\seti_boinc_2k3_2.2b-ben-joe\db\xml_util.cpp(663) : warning C4717: 'xml_match_tag': Rekursiv für alle Steuerelementpfade. Die Funktion verursacht einen Stapelüberlauf zur Laufzeit.
c:\i\sc\seti\seti_boinc_2k3_2.2b-ben-joe\db\xml_util.cpp(689) : warning C4717: 'xml_find_tag': Rekursiv für alle Steuerelementpfade. Die Funktion verursacht einen Stapelüberlauf zur Laufzeit.
Kompilieren...
------------------------------------------------------------------------------------------------------------------------
and here is the code ---->
// return true if the tag appears in the line
//
bool xml_match_tag(char* buf, char* tag) {
...
}
bool xml_match_tag(const std::string &s, char* tag)
{
return xml_match_tag(s.c_str(),tag);
663 } <----- here
size_t xml_find_tag( char* buf, char* tag) {
...
}
std::string::size_type xml_find_tag(const std::string &s, char* tag)
{
std::string::size_type p=xml_find_tag(s.c_str(),tag);
return (p!=strlen(s.c_str()))?p:(std::string::npos);
689 } <------- here
----------------------------------------------------------------------------------------------------------
have anybody of you a idea to prevent this
your suggestions are welcome....
regards heinz
Hi all,Last week, after a month of waiting I got the monitor. I ordered a silver colored monitor and when I opened the package I get a black one. I installed it, but it does not run properly( letters total blurred and fuzzy, in a distance of all 3cm). It was a desaster. So I sent these damaged hardware back to the shop and ordered back my money. All this trouble takes me a lot of time. Now I installed a 10 years old MAG CRT Monitor MX17S, it works fine and the letters are clear and exact written on it. Have some old hardware lying around in my working area, so this was possible.
I use a dual monitor installation for working. My second Monitor died last night. :'(
Today I ordered a new 22" multisync flatscreen from LG. Expected in ca. 4 days.
I have a problem with the CPU usage at the MMX dual machine.What is wrong, you think ? The pics show a normal distribution of CPU-time and usage for a dual CPU configuration.
...
Any hints are welcome.
Since I´m back it is so calm here in the forums.
Or are running all important informations behind the backstage ?
Sometime it seems and I have the feeling nobody is interested in further development of the s@h app.
Sure, a lot is already done. All thanks to the development crew.
-----------------------------------------------------------------------------------------
I asked in the Pre-Release -->Windows forum (22.09.07)
Please let me know if anybody of you are working with me together in further optimization of the source code.
More about 40 read it, but nothing. No answer till now........I can´t believe it.
-------------------------------------------------------------------------------------------------------------
:( :'(
Since I´m back it is so calm here in the forums.
Or are running all important informations behind the backstage ?
Sometime it seems and I have the feeling nobody is interested in further development of the s@h app.
Sure, a lot is already done. All thanks to the development crew.
-----------------------------------------------------------------------------------------
I asked in the Pre-Release -->Windows forum (22.09.07)
Please let me know if anybody of you are working with me together in further optimization of the source code.
More about 40 read it, but nothing. No answer till now........I can´t believe it.
-------------------------------------------------------------------------------------------------------------
:( :'(
// Part 1 allocation and init
// Part 2 generate fft coefficients, save into wisdom
// Part 3 generate chirp/fft pairs, do different calcs in preparation analyze
// Part 4 loop through chirp/fft pairs - this is the top level analysis loop.
// Part 4.1 chirping data
// Part 4.2 do transpose if needed
// Part 4.3 process data
// Part 4.4 analyze power over time (POT), set checkpoint
// Part 5. return the "best of" signals and do the rest
#ifndef USE_FFTW // FFTW now uses out of place transforms.
memcpy( WorkData, &ChirpedData[CurrentSub], int(fftlen * sizeof(sah_complex)) );
#endif
LOL, Here's one in seti_analyze that disappears if going to FFTW,
Code: [Select]#ifndef USE_FFTW // FFTW now uses out of place transforms.
memcpy( WorkData, &ChirpedData[CurrentSub], int(fftlen * sizeof(sah_complex)) );
#endif
I see a few of those.
Another thought. Has anyone attempted to use that FFTW codelet generator given that only a small portion of fftw is used? I have played with OCAML before, didn't seem hard.[but it was long enough ago to have forgotten everything :D]
Jason
...
#ifndef USE_FFTW // FFTW now uses out of place transforms.
// Commenting out the mempy()
// memcpy( WorkData, &ChirpedData[CurrentSub], int(fftlen * sizeof(sah_complex)) );
#endif
...
// Now fix the source for out of place IPP call properly
#if defined( USE_IPP )
ippsFFTInv_CToC_32fc(
// ( Ipp32fc * ) WorkData, // changing from this source
( Ipp32fc * ) &ChirpedData[CurrentSub], // to direct source for out of place
( Ipp32fc * ) WorkData, // leave as same destination
FftSpec[FftNum],
FftBuf );
... The reversibility of a complete FFT is only needed during baseline smoothing.
Joe
...
What do you think about this in benchmark.cpp ?
regards heinz
the yellow line was there....and still "undefined" variable WorkArea?, that is wierd :o
yes very different linenumbers, I give the analyzeFuncs.cpp a new structure......Ahh that's right, the improved model you showed me ... that's some good stuff mmm.
I will have a look at benchmark again.... make a trigger for the case we use a zero fill.
heinz
hmm... maybe Alex Kan or Joe has a good answer
questions I have are:
- Why benchmark an array of zeroes ?
- If zeroed array needs to be benched , why not test it 'fully' out of place (separate src/dest buffer like below)?
hmm... maybe Alex Kan or Joe has a good answer
questions I have are:
- Why benchmark an array of zeroes ?
- If zeroed array needs to be benched , why not test it 'fully' out of place (separate src/dest buffer like below)?
I don't know why Ben Herndon used the out of place form of parameters in the ippsFFTInv_CToC_32fc() calls, but he may have checked the actual code produced and determined that was slightly more efficient.I wracked my brain about this, and ultimately came to a similar (though more convoluted and speculative) conclusion. It would make sense to me if an explicit out of place call could make better use of the prefetch, cache and paralellism mechanisms we have discussed in a different context. An explicit in place call could not, (so far as I can see for now, through read write dependancies).
Joe
ahah I see.... now that IPP call is "In Place" You can do this:if we do this we get a error message ---->
...
if(pre_test == zero_out)
{
ippsFFTInv_CToC_32fc(
// ( Ipp32fc * ) out_buf, // Commented out this to make it inplace
( Ipp32fc * ) out_buf, // This is both source and destination
FftSpec,
NULL );
}
so it compiles
heinz
Surprise Surprise, a QxN build is faster on my Northwood :Phave a Northwood too --->
LOL
4- Chirp function Block Prefetch, memcpy++ zerocase & 3phase chirp Generic x86 Untested ~?.?%
measure its the best to try code and find optimal variants. ;D
the loop construct in pulsefind.cpp is ready now, but not measured.
Today I will squeeze the case-construct code.
have still some good ideas to eleminate code else and there...we will see...
have a strong modified chirpfft.cpp which we can try too
I am running vtune on the chirp one now to look for any p4 specific slowdowns, wickedly fast code though :D
have a strong modified chirpfft.cpp which we can try too
access of 23 000 .... I didn´t expected it..... looks like a hot thread
greatings to all who are reding here
heinz ;)
Hi Jason,
have you access to the "Pre-Release Applications" Forum ? If not, ask the moderators to give you access rights.
heinz
Joe and I are Mods, though this is more like the Maytag repairman's job around here.Hi Jason,
have you access to the "Pre-Release Applications" Forum ? If not, ask the moderators to give you access rights.
heinz
No, I didn't know there was such a place ::). Have you let the cat out of the bag again Heinz? Who are the Moderators? (Never did pay much attention to the tags on the posts... I guess I should really know that :D)
[Is there excellent stuff worth pestering a mod about in there I might want or need ?]
Thanks, Jason
Joe and I are Mods, though this is more like the Maytag repairman's job around here.thanks Gecko_R7! I have all the time in the world, no hurry. I have been informed by PM that some interesting/important things ARE in there, sounds good. Keep up with the intense moderation work :D
I think this can only be done by Admin rights (Simon), but have asked Joe if he is able to do it by chance as I don't see the option. Hang tight.
Joe and I are Mods, though this is more like the Maytag repairman's job around here.thanks Gecko_R7! I have all the time in the world, no hurry. I have been informed by PM that some interesting/important things ARE in there, sounds good. Keep up with the intense moderation work :D
I think this can only be done by Admin rights (Simon), but have asked Joe if he is able to do it by chance as I don't see the option. Hang tight.
Jason
Hi,
I am new to this form and would like to know which app is best for my Computer.
The OS is Windows 2000 pro Service Pack 4 the Cpu is Dual Xeon 2.0 GHZ which app would be the best for my setup!
Regards
ScanMan
Working now on a vectorized version of chirpfft.cppHi Heinz, I'm now on holidays :D, Are you looking at this one? I am trying to get reoriented after finishing study/work for the year, and am recovering after some serious celebrations :D. It's time to catch up!
heinz ;D
Heinz, keep in mind i did pay Aus$197 for my Wolfdale , which is really a piece of sand about 0.25 cm squared. That is more expensive, yet more useful than gold...197.00 Australian dollars = 118.33 Euros
Sometime I'm not sure what todo, SkullTrail yes/no with E5405 (http://www.intel.com/cd/products/services/emea/deu/processors/xeon5000/344535.htm) (178 Euro)Honestly Heinz, I'd say it'd be difficult to go past the Q6600 at the moment. I'd guess the Yorkfields are being held off 'till the stock of those clears a bit. Then the Yorkfields will be awesome [If this wolfdale is anything to go by]. I am getting the feeling that the Nehalem architecture will be a fairly radical departure from what we're used to, and it may take some time for the software to follow. Perhaps something like the OpenMP standard gives some insight there, many cores with shared memory.
or better wait for Nehalem (http://www.intel.com/technology/architecture-silicon/next-gen/index.htm?iid=tech_arch+body_45nm_nehalem)
or meanwhile a cheaper resulution: board XFX GeForce 7150/MCP630i (70 Euro) no graphiccard necessary, with a Intel Core2 Quad Q9450 4x2.67GHz BOX (300 Euro), case ram disk..all together ca 680 Euro for the hardware + software XP Professional (130 Euro) for testing our parallel stuff.....
Sometime I'm not sure what todo, SkullTrail yes/no with E5405 (http://www.intel.com/cd/products/services/emea/deu/processors/xeon5000/344535.htm) (178 Euro)Honestly Heinz, I'd say it'd be difficult to go past the Q6600 at the moment. I'd guess the Yorkfields are being held off 'till the stock of those clears a bit. Then the Yorkfields will be awesome [If this wolfdale is anything to go by]. I am getting the feeling that the Nehalem architecture will be a fairly radical departure from what we're used to, and it may take some time for the software to follow. Perhaps something like the OpenMP standard gives some insight there, many cores with shared memory.
or better wait for Nehalem (http://www.intel.com/technology/architecture-silicon/next-gen/index.htm?iid=tech_arch+body_45nm_nehalem)
or meanwhile a cheaper resulution: board XFX GeForce 7150/MCP630i (70 Euro) no graphiccard necessary, with a Intel Core2 Quad Q9450 4x2.67GHz BOX (300 Euro), case ram disk..all together ca 680 Euro for the hardware + software XP Professional (130 Euro) for testing our parallel stuff.....
Jason
Likely to be massive price gouging and very little supply initially...ltd mobo options, and buggy release bios's. Wouldn't be surprised for it to be at least Q2 09' before we see decent pricing and availability for us mere mortals that have budgets to consider.
We should also see some nice price drops on Penryn/Yorkfield and perhaps a new stepping as Nehalem is released.
Well priorities have a way of shifting depending on need. As you point out p4's & AMDs of SSE2 vintage are still extremely popular according to boincstats, and dominate throughput in many respects.
What the tests seem to be showing is that Alex pretty well nailed the Core2 code, and unless we decide to tackle the other end there may be little left to do there for now (Unless, that is, some of the relaxed validation requirements that have been spoken about are put in place, then the parallelism race may be back on in force).
Early p4's have special characteristics to do with cache that aren't necessarily all that happy with techniques used in builds targeted for the core2 architecture. There are speed improvement showing in the p4(SSE3) I tested, but not as great as the core2 improvements. There might be plenty of room to tweak that and the SSE3 instructions may as well be macro encapsualted while we're there, allowing SSE2 substitution.
There is though possibly still quite a bit more opportunity to squeeze more performance from the core2 build first. We have spoken about profile guided optimisations, which haven't been touched yet, and in fact no profiles have even been run yet to identify possible bottlenecks or problems with the build, That is why, in my book, it is still considered pre-alpha. Valid results is one thing, but releasing substandard builds I'd rather leave to the software companies who have the excuse of pressure from the marketing department.
Jason
The new servers are commingyeah, the prices of FB-Dimms or DDR3 too is why I have stayed with a DDR2 desktop mainboard. It is a nice step up for you and I think there is enough processing power there to decommission your other machines, then you will quickly save enough money in electricity and the prices of RAM will drop far enough then you can put 4 Gig in easily ;D
Altos G540 TT.G54E0.033
Altos G540 Xeon QC E5405 2.0GHz/1333 MHz/12MB, 2x1GB FBD 667, 1xSAS/SATA HDD cage (4 bays, second cage optional), SATA 6 ch on board, no HDD and Carrier, DVD, Dual Gigabit Lan Onboard, PS2 Maus, ohne Keyboard, EasyBuild 7.1, 3 Jahre Garantie und Vor-Ort-Service Next
1321,18 Euro there (http://www.f-m-shop.de/seiten/frame_ga_schotte.cfm?kat=gArtikel&nav=2&artnr=991388505)
2GB RAM looks a little be smal for a quad, but the price is hot ::)
heinz
Today I made some fotos of the powersupply. Have a look at Xeon V8-Server (http://www.britta-d.de/bilder/server/server.htm).So are you saying you can have only one Xeon sticker on the front ? LOL, maybe if you put it there Francois will come and make you peel it off again ... ;D If he does this I insist that you kick his bum!
There are 4 connectors for PCi-E (for the possible 4 graphic cards of the D5400XS)
With the processors I got the Xeon label, but it is not allowed me to place it on the front bezel of the case without a extra licence of Intel, huuuh :o
heinz
No, in every cpu package is a Xeon-inside label (so I have 2 of them), but I did not open the second package till now. Show Xeon inside (http://www.britta-d.de/bilder/server/page11.htm) and read the full text.LOL, I like that .. "we give you a sticker, but you can't use it unless you agree to sign an agreement, and by doing so your are agreeing not to use the sticker on any system that we don't want you to!" LOL ::)
.....Even with VM I cant emulate SSE4.1 on my Venice IMHO ;)LOL, maybe if you clench really hard like Hiro from Heroes it will magically turn into a Wolfdale ;D. Although I use VM's for testing, I think any kind of timing benchmarks would be a waste of time. This is why I wish to run dual boot to test 64 bit for myself, might as well get you some sse41 PGO data while I'm there, running native.
jason,I think your new server is designed with something like this in mind :
as I have seen hardware virtualisation is available there. ;D
Today I got the email from Redmont for the evaluation with a lot of further infos.
thanks Microsoft. ;D
for all freaks of CF
new 300x CF are available:
4096MB Transcend Compact Flash Card 300x ca. 45,00 Euro
8192MB Transcend Compact Flash Card 300x ca. 82,50 Euro
16GB Transcend Compact Flash Card 300x ca. 173,30 Euro
Speed 300X CompactFlash® Cards running Dual-Channel Ultra DMA Modus 5*
- Hardware ECC (Error Correction Code) Fehlerkorrektur
- Support IDE PIO Modus 6 and Ultra DMA Modus 5
------------------------------------------------------------------------
to use it you need a adapter (http://www.britta-d.de/bilder/server/page25.htm)
;D
you want to know some more about the GRID.. I invite you to have a look at the Grid Café (http://gridcafe.web.cern.ch/gridcafe/)Welcome to the EGI (http://web.eu-egi.eu/)
Greetings! :) Keep up !looks like I can OC the machine up to 4GHz
Hi Heinz,I have seen this:
Looks like you can cook some eggs for breakfast on your RAM ;D I guess as you push the CPU's a bit more this may be the first thing that may require some compromise (perhaps on the 1:1 ratio). If you drop the voltage on those a little do you get errors?, just curious
Jason
New approach to grid computing ;)Thank you very much to remember me, I have got a email too ;)
http://www.codeproject.com/KB/silverlight/gridcomputing.aspx
ey Heinz! time to drag out the development documentation again ;DToday I installed TortoiseSVN, latest Version. If I have done all the necessary adjustments then I will open the development documentation "H_Performance_Doku". It consists of 1292 files, summary 187 MB.
News from the V8-Xeon:Today fedora9 runs several hundred of updates, although I installed the latest stable version.
Fedora9_x86_64 is installed now on the SKT parallel to Vista64 on the second disk.
heinz
Hi Heinz,The only thing is the sensor of the case, which i placed between the 2 FB-DIMMs, in the last 5 minutes this sensor goes down from 71 to 65°C, after I ended the measuruing with the programms Everest, CPUID-Hardwaremonitor and Processexplorer.
That seems rather incredible. Have you got some kind of thermometer to put it there so you can see if that is anywhere near accurate?
Heinz, 2.2 is too old. Though I can't read the German errors I would avoid older versions (that old) like the plague because of Boincapi, compiler, and target platform changes. please consider starting with AK_v8 build as we tried to make it much easier. simply changing the ICC & IPP directories (in the project includes and linker linraiy paths) should allow you to build against our boincapi. If you need some help let me know, as I am coming to some heavy multithreaded opportuntities and would like you to be up to speed for some collaboration.on my old xp-machine this older version 2.2 compiles complete without error. So I thought it is a good complete test.
Jason
REM /X64 - target Windows for the X64 bit platform
I cant believe it, Vista 64 shows PROCESSOR_ARCHITECTURE AMD64 :o
heinz
Good news! :) But price still high...And Motherboards are NOT still stable!
Does it need DDR3 too or some boards with DDR2 support exist?
What ultimate upgrade price (all parts that should be replaced) will be?
@
You are among the first to receive notification of the groundbreaking Intel® Parallel Composer Beta. Download this exciting new tool and get instant access to an advanced parallelism C/C++ compiler, debugger, and libraries that can change the way you develop parallel applications. @
;D
Will look...
@Heinz: Do you happen to have any single and multithreaded FFT processing times benched on your skulltrail? Time for 1,2,4 & 8 threads would be nice for 32k element &/or 128k elements, if you have them.Hi Jason,
I'm trying to verify/refine some efficiency calculations & have no reference but my dual core.
Jason
@I'm registered and downloading now..
You are among the first to receive notification of the groundbreaking Intel® Parallel Composer Beta. Download this exciting new tool and get instant access to an advanced parallelism C/C++ compiler, debugger, and libraries that can change the way you develop parallel applications. @
;D
Will look...
Just Released: AMD Core Math Library v4.2.0Out of my view we can link from different libs....often this is used in scientific work.
OMG it's FORTRAN library %)
http://developer.amd.com/cpu/Libraries/acml/downloads/Pages/default.aspx#downloads
And it has many flavors... Interesting, can it be used w/o any FORTRAN installation, just as simple lib-file? ....
@Heinz: Do you happen to have any single and multithreaded FFT processing times benched on your skulltrail? Time for 1,2,4 & 8 threads would be nice for 32k element &/or 128k elements, if you have them.compiled the fftw project (single thread) as 32 bit
I'm trying to verify/refine some efficiency calculations & have no reference but my dual core.
Jason
Thanks Heinz,CPU speed 2398 MHz
Could you let me know:
- Current CPU speed at time of test
- Cache sizes per package
- Bus speed
Thanks Heinz,CPU speed 2398 MHz
Could you let me know:
- Current CPU speed at time of test
- Cache sizes per package
- Bus speed
FSB speed 400(QP) 1600
Cache sizes per package ... I must look up ( where can I find in the source ? )
ahh.. cpu package.. 12 MB
What was your original bench? non-sse floats fftw 3.1.2?Configuration: Active(Release float SSE) Platform: Active(Win32)
Threads number is 5test before that is 4, next is 16, so it seems a bit weird. I have some fibonnaci example project I did here with TBB a while ago. Will see if I can dig it out.
Hi Heinz, Why did it choose 5 threads instead of 8 ?5:8 should be 5(1) to 8, hmmm 1:4 did work as you can seeQuoteThreads number is 5test before that is 4, next is 16, so it seems a bit weird. I have some fibonnaci example project I did here with TBB a while ago. Will see if I can dig it out.
looked up.. this works correct with one lineHi Heinz, Why did it choose 5 threads instead of 8 ?5:8 should be 5(1) to 8, hmmm 1:4 did work as you can seeQuoteThreads number is 5test before that is 4, next is 16, so it seems a bit weird. I have some fibonnaci example project I did here with TBB a while ago. Will see if I can dig it out.
echo off
echo please wait
echo fibonacci result in fibonacciopt_1000_out.txt
rem no second parameter means standard(1:4)
fibonacci.exe 1000 >fibonacciopt_1000_out.txt
echo fibonacci 1000 5:8
fibonacci.exe 1000 5:8 >>fibonacciopt_1000_out.txt
echo fibonacci 1000 16
fibonacci.exe 1000 16 >>fibonacciopt_1000_out.txt
echo fibonacci 1000 32
fibonacci.exe 1000 32 >>fibonacciopt_1000_out.txt
echo fibonacci 1000 64
fibonacci.exe 1000 64 >>fibonacciopt_1000_out.txt
echo ready
you can compare with the other result file to see differences
heinz
Intel(R) C++ Compiler 10.1 Integration(s) in Microsoft Visual Studio* is already installed.
Installation can continue; however, you will not be able to use the Intel C++ Compiler 10.1 or 9.0 within the Visual Studio IDE
Heinz, have 64 & 32 bit Parallel composer beta (update 2) installed .... Where can I find the fibonacci sample? (the stuff I see here looks more boring)fibonacci is part of TBB
ahh.. cpu package.. 12 MBThis is wrong, unless you have a not-yet-released version of the Nehalem processor.
Hmmm, I have TBB on my other (32-bit) drive ... maybe I can install it here, will try.I did upload the fibonacci project to our testproject
[Hmmm, so parallel composer doesn't actually have TBB in it then .... ??? that seems a bit odd, maybe they expect you'll use pure openmp.. what about IPP, I suppose that's not there either which would make this ICC 11 ?]
32 & 64 bit fibonacci sample built & ran. Will consider fully migrating to 64 bit platform for holidays in a few weeks. It'll be painful, but about time probably.
Jason
Hi Leaps! .. Nahhh .. It's 2xXeon Quads on a Skultrail Mobo :) [Heinz, Please check cache size with CPU-Z]there you can see CPUID (http://www.britta-d.de/bilder/server_oc/page9.htm)
Ahhh, 6 meg per package ( 1.5 meg per core )... Okay, yep it is 12 meg total for the 8 cores.
Compared 32 bit ICC 10.1 / TBB 2.0 build of fibonacci, and it IS slower than Parallel composer 32 bit build under XP64 ... Will have to try that build under XP32 to confiirm though. I will probably update all my ICC/IPP base packages as soon as I get time, in a few week.
Jason
Threads number is 2
Shared serial (mutex) - in 0.286294 msec
Shared serial (spin_mutex) - in 0.196978 msec
Shared serial (queuing_mutex) - in 0.301214 msec
Shared serial (Conc.HashTable) - in 4.313505 msec
Parallel while+for/queue - in 1.485761 msec
Parallel pipe/queue - in 1.980293 msec
Parallel reduce - in 0.523162 msec
Parallel scan - in 0.338611 msec
Parallel tasks - in 0.566134 msec
Threads number is 2
Shared serial (mutex) - in 0.279819 msec
Shared serial (spin_mutex) - in 0.208223 msec
Shared serial (queuing_mutex) - in 0.284642 msec
Shared serial (Conc.HashTable) - in 4.461598 msec
Parallel while+for/queue - in 1.718736 msec
Parallel pipe/queue - in 2.188073 msec
Parallel reduce - in 0.571781 msec
Parallel scan - in 0.357319 msec
Parallel tasks - in 0.534837 msec
Threads number is 3
Shared serial (mutex) - in 162.014407 msec
Shared serial (spin_mutex) - in 11.609819 msec
Shared serial (queuing_mutex) - in 50.960339 msec
Shared serial (Conc.HashTable) - in 401.327768 msec
Parallel while+for/queue - in 93.399315 msec
Parallel pipe/queue - in 164.994829 msec
Parallel reduce - in 27.500117 msec
Parallel scan - in 22.918168 msec
Parallel tasks - in 25.904447 msec
Threads number is 3
Shared serial (mutex) - in 76.449678 msec
Shared serial (spin_mutex) - in 13.449323 msec
Shared serial (queuing_mutex) - in 50.961819 msec
Shared serial (Conc.HashTable) - in 413.186277 msec
Parallel while+for/queue - in 93.995606 msec
Parallel pipe/queue - in 171.541281 msec
Parallel reduce - in 28.647254 msec
Parallel scan - in 27.231642 msec
Parallel tasks - in 24.389762 msec
No, just used default which was 100... will try 1000
[Later:] Fastest 32 bit run built on XP32 ICC10.1 / TBB2.0 now 3 threads :o:QuoteThreads number is 3Now you know why I choosed 5 .. a not even number
We can create every number of threads 1, 2, 3, 4.. 128, 256, 512 etc. not even numbers also.
and we can use /QxHOST ---> Best performance on latest features of the processor supported by the compilation host.
::)
heinz
sample above compiled with MSC-Compiler@Heinz: Do you happen to have any single and multithreaded FFT processing times benched on your skulltrail? Time for 1,2,4 & 8 threads would be nice for 32k element &/or 128k elements, if you have them.compiled the fftw project (single thread) as 32 bit
I'm trying to verify/refine some efficiency calculations & have no reference but my dual core.
Jason
/I "." /I ".." /I "../libbench2" /I "../api" /I "../kernel" /I "../dft" /I "../rdft" /D "WIN32" /D "NDEBUG" /D "_CONSOLE" /D "FFTW_SINGLE" /D "BENCHFFT_SINGLE" /D "HAVE_SSE" /D "_VC80_UPGRADE=0x0710" /D "_MBCS" /FD /EHsc /MT /Fp".\bench___Win32_Release_float/bench.pch" /Fo".\bench___Win32_Release_float/" /Fd".\bench___Win32_Release_float/" /W3 /nologo /c /errorReport:prompt
Results:
C:\Windows\system32>echo off
fftw-3.1.2 benchfsse(VS2005) started
benchf_sse.exe -opatient 8 16 32 64 128 256 512 1024 2048 4096 8192 16384 32768
131072
Problem: 8, setup: 300.32 us, time: 169.69 ns, ``mflops'': 707.16
Problem: 16, setup: 288.86 us, time: 332.84 ns, ``mflops'': 961.43
Problem: 32, setup: 7.91 ms, time: 726.79 ns, ``mflops'': 1100.7
Problem: 64, setup: 27.46 ms, time: 1.67 us, ``mflops'': 1148.4
Problem: 128, setup: 62.98 ms, time: 4.19 us, ``mflops'': 1069.1
Problem: 256, setup: 137.48 ms, time: 9.18 us, ``mflops'': 1115
Problem: 512, setup: 267.80 ms, time: 20.95 us, ``mflops'': 1099.6
Problem: 1024, setup: 575.47 ms, time: 46.10 us, ``mflops'': 1110.7
Problem: 2048, setup: 1.37 s, time: 99.17 us, ``mflops'': 1135.8
Problem: 4096, setup: 3.42 s, time: 220.42 us, ``mflops'': 1115
Problem: 8192, setup: 8.83 s, time: 530.79 us, ``mflops'': 1003.2
Problem: 16384, setup: 21.99 s, time: 1.13 ms, ``mflops'': 1014.9
Problem: 32768, setup: 53.80 s, time: 2.41 ms, ``mflops'': 1020
Problem: 131072, setup: 369.12 s, time: 9.89 ms, ``mflops'': 1126
fftw-3.1.2 benchfsse ended.
Drücken Sie eine beliebige Taste . . .
----------------------------------------------------------------------------------------------------
For the threaded variants I must first read doku again...
Did you mean this ? or if you want some other Compiler options let me know..
If I have installed the Intel® Parallel Composer Beta, I will recompile the project...
regards heinz
sample above compiled with MSC-Compiler@Heinz: Do you happen to have any single and multithreaded FFT processing times benched on your skulltrail? Time for 1,2,4 & 8 threads would be nice for 32k element &/or 128k elements, if you have them.compiled the fftw project (single thread) as 32 bit
I'm trying to verify/refine some efficiency calculations & have no reference but my dual core.
Jason
/I "." /I ".." /I "../libbench2" /I "../api" /I "../kernel" /I "../dft" /I "../rdft" /D "WIN32" /D "NDEBUG" /D "_CONSOLE" /D "FFTW_SINGLE" /D "BENCHFFT_SINGLE" /D "HAVE_SSE" /D "_VC80_UPGRADE=0x0710" /D "_MBCS" /FD /EHsc /MT /Fp".\bench___Win32_Release_float/bench.pch" /Fo".\bench___Win32_Release_float/" /Fd".\bench___Win32_Release_float/" /W3 /nologo /c /errorReport:prompt
Results:
C:\Windows\system32>echo off
fftw-3.1.2 benchfsse(VS2005) started
benchf_sse.exe -opatient 8 16 32 64 128 256 512 1024 2048 4096 8192 16384 32768
131072
Problem: 8, setup: 300.32 us, time: 169.69 ns, ``mflops'': 707.16
Problem: 16, setup: 288.86 us, time: 332.84 ns, ``mflops'': 961.43
Problem: 32, setup: 7.91 ms, time: 726.79 ns, ``mflops'': 1100.7
Problem: 64, setup: 27.46 ms, time: 1.67 us, ``mflops'': 1148.4
Problem: 128, setup: 62.98 ms, time: 4.19 us, ``mflops'': 1069.1
Problem: 256, setup: 137.48 ms, time: 9.18 us, ``mflops'': 1115
Problem: 512, setup: 267.80 ms, time: 20.95 us, ``mflops'': 1099.6
Problem: 1024, setup: 575.47 ms, time: 46.10 us, ``mflops'': 1110.7
Problem: 2048, setup: 1.37 s, time: 99.17 us, ``mflops'': 1135.8
Problem: 4096, setup: 3.42 s, time: 220.42 us, ``mflops'': 1115
Problem: 8192, setup: 8.83 s, time: 530.79 us, ``mflops'': 1003.2
Problem: 16384, setup: 21.99 s, time: 1.13 ms, ``mflops'': 1014.9
Problem: 32768, setup: 53.80 s, time: 2.41 ms, ``mflops'': 1020
Problem: 131072, setup: 369.12 s, time: 9.89 ms, ``mflops'': 1126
fftw-3.1.2 benchfsse ended.
Drücken Sie eine beliebige Taste . . .
----------------------------------------------------------------------------------------------------
For the threaded variants I must first read doku again...
Did you mean this ? or if you want some other Compiler options let me know..
If I have installed the Intel® Parallel Composer Beta, I will recompile the project...
regards heinz
C:\Windows\system32>echo off
compiled with Parallel Composer Configuration(Release float SSE) Platform(Win32)
fftw-3.1.2 benchf_sse started
benchf_sse.exe -opatient 8 16 32 64 128 256 512 1024 2048 4096 8192 16384 32768
131072
Problem: 8, setup: 241.93 us, time: 49.93 ns, ``mflops'': 2403.6
Problem: 16, setup: 276.57 us, time: 94.39 ns, ``mflops'': 3390
Problem: 32, setup: 7.91 ms, time: 117.86 ns, ``mflops'': 6787.9
Problem: 64, setup: 26.76 ms, time: 219.35 ns, ``mflops'': 8753.3
Problem: 128, setup: 61.71 ms, time: 447.42 ns, ``mflops'': 10013
Problem: 256, setup: 124.16 ms, time: 855.56 ns, ``mflops'': 11969
Problem: 512, setup: 238.18 ms, time: 1.99 us, ``mflops'': 11575
Problem: 1024, setup: 403.56 ms, time: 4.47 us, ``mflops'': 11455
Problem: 2048, setup: 719.56 ms, time: 10.62 us, ``mflops'': 10611
Problem: 4096, setup: 1.41 s, time: 25.84 us, ``mflops'': 9510.4
Problem: 8192, setup: 3.14 s, time: 58.67 us, ``mflops'': 9076.4
Problem: 16384, setup: 7.01 s, time: 125.16 us, ``mflops'': 9163.6
Problem: 32768, setup: 16.08 s, time: 279.92 us, ``mflops'': 8779.5
Problem: 131072, setup: 87.35 s, time: 1.29 ms, ``mflops'': 8658.3
fftw-3.1.2 benchf_sse ended.
with 128K 8658,3 mflops
best relation ~1:10
let's everybody make his own thoughts..
heinz
you gotta be carefull with fftw and which compiler to use. From my own experience the pre-packaged gcc builds where always faster than the icc compiled code !
since yesterday the runtime from ap5.03 (ap_5.03r112_SSE3.exe) decrease from 15 hours to 10 hours.
in all these shorter wu's I see now: repetitive pulses: 30, percent blanked: 0.00
looks like the wu finished in this case.
hmmm....anything happened on the serverside ?
on 4th april I calculated 19 wu's, all finished around 10 hours.
have a look at my machine http://setiathome.berkeley.edu/result.php?resultid=1196217010
heinz
What wonders me more is that i have seen a similar course of RAC (http://www.echtbaer.de/Images/RAC_P9500_090427.png) on my P9500 about a week ago.
My RAC drop down like a rock. (http://www.britta-d.de/images/astropulse/rac_down.jpg)
18th april rac=17500
26th april rac=15400
Although crunching 24/7 astropulse 5.03 all the time this surprized me.
Usually I run 12 wu/per day, sometimes some more, depending on the blanking of the wu. Looks like I get a wu-series with high blanking now.
My pendings increased from 146,000 up to
Pending credit: 174,653.22
----------------------------------------
have a look at the statistics (http://boincstats.com/stats/boinc_host_graph.php?pr=bo&id=f92199c6a274d674b15745b604c432e2)
Tonight I will run the cache dry to make some OC experiments and reconfigurations.
heinz
Today I will go to my neighbour and try to reach the VMs over internet from there.have had some problems todo it
problems resolved now, Jason can work on my VM's now ;DToday I will go to my neighbour and try to reach the VMs over internet from there.have had some problems todo it
Tried the Beta. Hated the removal of the compiler switches I like, then went back to ICC. They've got basically the same stuff, repackaged IMO, but targeted to a different target customer.
Get my W7 Ultimate yesterday ;DHave fun ! :D
Downgraded to Catalyst 9.2.All discussion of this app in beta area still.
Downloaded ap_5.05_win_x86_SSE3_BROOK_r280 and like to start testing.
Has someone a working app_info.xml for this purpose?
Furthermore a read something about a BOINC Beta account. Do I need that? Is it different from my 'normal' BOINC account?
This morning I found the V8-Xeon down..I'm still waiting for the power supply :'(
The 1000W AXP Power supply died....phhhh :'(
Fighting now with RMA...
New power supply has arrived now. This time I got a SuperFlower Crystal Plus 1000W.Good news,
My old one AXP 1000W (http://www.britta-d.de/bilder/server/page14.htm) was not available, so I got the Superflower (http://www.super-flower.de/index.php?id=46).
Tomorrow we will see if this woks with my D5400XS.
Intel(R) Atom(TM) CPU 230 @ 1.60GHz [x86 Family 6 Model 28 Stepping 2]
Number of processors 2
Coprocessors NVIDIA ION (256MB) driver: 19107
Operating System Microsoft Windows Vista
Home Premium x86 Edition, Service Pack 2, (06.00.6002.00)
Memory 1790.45 MB
Cache 512 KB
Swap space 3831.9 MB
Total disk space 53.74 GB
Free Disk Space 26.28 GB
Measured floating point speed 678.2 million ops/sec
Measured integer speed 1630.76 million ops/sec
Would also be great to see what a 30 day diet of Cuda processing produces.
Would also be great to see what a 30 day diet of Cuda processing produces.
Today a month is over, 2 days the machine was used to play, install software and show videos...
First seen on 2009-11-08 06:38:13
CPU Intel(R) Atom(tm) CPU 230 @ 1.60GHz
Number of CPU's (number of (virtual) cores) 1(2)
Operating System and version Microsoft Windows Vista
Current Credit (based on incremental update) 51,391.52
BOINC World position based on credit (based on incremental update) 534,061
Recent average credit RAC (projects accumulated) 1,901.33990
Recent average credit RAC (according to BOINCstats) 1,554.07922
Average credit per CPU second 0.027428
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
here is the full statistic: Host ID "6187800" (http://boincstats.com/stats/boinc_host_graph.php?pr=bo&id=2c331715ddea0590c336066226571b03)
summary we can say it crunches ~50 000 credits per month
and get a RAC of 1900 - 2000 depending on project
have fun ;D
if you have FB-DIMMS on your server and you have some good ideas to cool them down, so let me know.looking around I found in inquirer FB-DIMM with Heatpipe cooling (http://www.theinquirer.net/inquirer/news/1009915/first-inqpressions-kingston) a very interesting review.
regards ;D
>10C difference between edge memory slots?! Too big gradient IMO to call this cooling system good one.you are right, its more a funny kit than a good cooling solution. :(
OK, I will run this EVO Cyclone cooler, but I must search for a better more effective cooling solution.
Lets everybody make his own conclusion. ;)
The silly question I have is the fan blowing Down onto the cooling fins or UP pulling case air through the fins...The fan in the tube is blowing the cool air from outside onto the FB-DIMMs. The fan does not blow on the cooling fins left and right. These cooling fins goes still under the fan along. When the fan blows the air on these fins under it, they cool down, but this has no big effect if the case temp is 40 grd celsius. (Therefore I must the case open now).
So looking at the pictures you have posted you have what looks like a dead space for air movement in the area of the RAM. What I think I would do is install one of the fans that covers a "blade slot" in the back of the case to exhaust from that area and insure that the RAM Fan is pulling outside air (forcing down onto to the cooling fins). Then you are creating specific airflow for just the RAM.
no work from seti on all my machines now.1>AP SSE3ATOM Win32 (Microsoft VC++ Environment)
running collatz and milkyway...
time to compile something for the ION platform :)
Strangely enough none of the Atom builds work proberly, allthough they are SSE3, and should be compatible.'Should be', though I believe ATOM has an extra instruction (MOVBE) which is available in our 45nM Core2's (at least)... So ATOM builds are really ATOM specific, though they should run on later Intels OK. (The SSE3 Qopt one uses the generic SSE3 options ... nice to know it works on Phenom II)
...
Thats OK, but then they should be marked (S)SSE3 or something to that effect, since they are not really SSE3 compatible.
still leaved damned Vista on my netbook so can't say it works fast, but Chrome is nice brower indeed. Started to use it on netbook firstly and now switched to it on my home desktop too :)
And for record: Atom is SSSE3-compatible CPU. In lacks x64 mode and SSE4.* only.
...Well, A few ideas on that run:
Strange that the ATOM switch shows "slower" on the DMH1023 WU. :-\
Not sure I trust that result.
...
Quick timetable
WU : ap_18se08aa_B6_P1_00046_1LC25.dat
ap_5.05r168_SSE3.exe : 2403.913 secs CPU
ap_5.05r303_ATOM_ICC_Qopt.exe : 2163.079 secs CPU
Speedup : 10.02%
Ratio : 1.11 x
ap_5.05r303_SSE3_ICC_Qopt.exe : 1919.093 secs CPU
Speedup : 20.17%
Ratio : 1.25 x
WU : DMH1023rr_ap_21oc08ab_B2_P0_00081_20081130_08605.dat
ap_5.05r168_SSE3.exe : 1952.649 secs CPU
ap_5.05r303_ATOM_ICC_Qopt.exe : 1671.145 secs CPU
Speedup : 14.42%
Ratio : 1.17 x
ap_5.05r303_SSE3_ICC_Qopt.exe : 1675.482 secs CPU
Speedup : 14.19%
Ratio : 1.17 x
WU : JasonMediumrr.dat
ap_5.05r168_SSE3.exe : 13857.850 secs CPU
ap_5.05r303_ATOM_ICC_Qopt.exe : 6451.858 secs CPU
Speedup : 53.44%
Ratio : 2.15 x
ap_5.05r303_SSE3_ICC_Qopt.exe : 6548.376 secs CPU
Speedup : 52.75%
Ratio : 2.12 x
WU : JasonShortrr.dat
ap_5.05r168_SSE3.exe : 3752.620 secs CPU
ap_5.05r303_ATOM_ICC_Qopt.exe : 3227.926 secs CPU
Speedup : 13.98%
Ratio : 1.16 x
ap_5.05r303_SSE3_ICC_Qopt.exe : 3236.210 secs CPU
Speedup : 13.76%
Ratio : 1.16 x
WU : Raistmer_tinyrr.dat
ap_5.05r168_SSE3.exe : 1186.544 secs CPU
ap_5.05r303_ATOM_ICC_Qopt.exe : 810.191 secs CPU
Speedup : 31.72%
Ratio : 1.46 x
ap_5.05r303_SSE3_ICC_Qopt.exe : 813.795 secs CPU
Speedup : 31.41%
Ratio : 1.46 x
WU : sigindrr.dat
ap_5.05r168_SSE3.exe : 5153.165 secs CPU
ap_5.05r303_ATOM_ICC_Qopt.exe : 4008.071 secs CPU
Speedup : 22.22%
Ratio : 1.29 x
ap_5.05r303_SSE3_ICC_Qopt.exe : 4072.968 secs CPU
Speedup : 20.96%
Ratio : 1.27 x
...It certainly can, and is probably the case here. Some platforms seem to converge quite quickly on wisdom, some take longer. I reckon it depends on how fftw arranged the heuristics in that initialisation, and to whether it finds the best codelet sequences soon or later in allowed time limits.
So, perhaps Wisdom gen time does have noticebale impact? :-\
...
For 10? I think you need two million. Oh, and I thought you were German. :)10 mio ? -->Target will be reached in 41.63 days on April 4 2010 (if i have no hardware outage)
Congrats anyway.
9 Mio total todayFor 10? I think you need two million. Oh, and I thought you were German. :)10 mio ? -->Target will be reached in 41.63 days on April 4 2010 (if i have no hardware outage)
Congrats anyway.
:)
Kommt hier eigentlich mal was Produktives, oder wird hier nur das "optimale" System präsentiert?natürlich gibts was productives, wir arbeiten an speed optimierten astropulse Apps
got the 10 Mio on 4th of April ;)9 Mio total todayFor 10? I think you need two million. Oh, and I thought you were German. :)10 mio ? -->Target will be reached in 41.63 days on April 4 2010 (if i have no hardware outage)
Congrats anyway.
:)
10 mio ? -->Target will be reached in 18.28 days on March 30 2010
;)
got the 10 Mio on 4th of April ;)
...
4h:02min, not bad for the ION chip,
so we know now that the app works on this chipset too.
regards Heinz
I'll soon be migrating also, gradually, to VS2008 for primary development, since the nVidia nSight stuff is made to work on that. Might make life a bit easier if we are on similar platforms/environment.Yeah, I'm on VS2008 too already :)
Yeah, I'm on VS2008 too already :)
@ Heinz,Hi Gecko,
Is it worth revisiting updated Atom CPU builds based on more recent source?
Even 5-10% differences are worthwhile when current Atom AP is @ 80 hours.
...You know I changed to W7 now, so a lot of updates are necessary....
Yes I have VS2008 installed already and latest Intel Compiler(065), now I have installed SVN but I cant find this config-file where I had todo the changes for the SVN-autoprops....You know I changed to W7 now, so a lot of updates are necessary....
I am moving to VS2008 w/sp1 Heinz (for Cuda development mostly) So I'd advise going to that config if you have the option.
Oh heinz, nearly forgot, sorry:I found it this way and made the necessary changes. Thank you very much. ;)
If you are having trouble locating the appication data folder on Win7, you can open a 'My Computer' windows explorer window, then type %APPDATA% in the address bar & press enter ... that should take you inside the mysterious hidden bowels of Windows quickly. Look around from there and you should find the right place (subversion folder) fairly easily.
Hm.... Atom is x86-only AFAIK...
Wow, good news! Then my next netbook could be x64 one ;D
The humble webcentric Netbook computer is getting a significant image boost this week following the official launch of Acer's Aspire One D255, which arrives as the very first Netbook to hit the U.S. market armed with a snazzy dual-core Intel Atom processor.
Sporting a typically pocket-friendly price of just $399 USD, the D255 comes equipped with the improved 1.5GHz Atom N550 platform (http://ark.intel.com/Product.aspx?id=50154&code=N550), 1GB of DDR3 RAM (1066MHz), a spacious 250GB hard drive (5400RPM), a 10.1-inch LCD screen with a resolution of 1024 x 600, and the Windows 7 Starter operating system.
Other contributing elements packed into Acer's latest ultra portable computer include integrated Intel GMA 3150 graphics, 802.11b/g/n Wi-Fi connectivity, 10/100 Ethernet, a trio of USB 2.0 ports, VGA out, a capable six-cell 4400mAh battery, a multi-card reader, and chassis colour choices of red, black, aquamarine or sandstone.
According to tech site Liliputing, the Aspire One D255 is presently only on sale through the Home Shopping Network, though we expect more widespread availability is likely just around the corner.
Device: GeForce GTX 480, 810 MHz clock, 1503 MB memory.
Compiled with CUDA 3000.
--------CUFFT------- ---FFT-------------- ---IFFT------
N Batch Gflop/s GB/s error Gflop/s GB/s error Gflop/s error
8 131072 7.1 7.6 1.4 140.0 149.4 1.1 140.5 1.1
16 65536 16.1 12.9 1.7 183.1 146.5 1.0 183.7 1.0
64 16384 259.2 138.2 1.4 280.0 149.4 1.4 279.7 1.4
256 4096 352.2 140.9 1.4 352.8 141.1 1.5 352.0 1.5
512 2048 413.3 146.9 1.8 411.8 146.4 1.8 412.2 1.8
Errors are supposed to be of order of 1 (ULPs).
Hi Heinz,Hi Jason,
Numbers come out different when you change to the same data set size that Multibeam apps use ( 1*1024*1024 complex data points).
CUFFT is not very fast at the small sizes for that small amount of data. It gets better relatively as the FFT size goes up. I haven't optimised these custom ones (So they remain ~G80 GPU arranged), but did change the results to give in-otder output. Didn't need two-way, so made forward & inverse transforms instead.
You can see CUFFT goes pretty slowly when doing many small transforms on our smaller dataset.Device: GeForce GTX 480, 810 MHz clock, 1503 MB memory.
Compiled with CUDA 3000.
--------CUFFT------- ---FFT-------------- ---IFFT------
N Batch Gflop/s GB/s error Gflop/s GB/s error Gflop/s error
8 131072 7.1 7.6 1.4 140.0 149.4 1.1 140.5 1.1
16 65536 16.1 12.9 1.7 183.1 146.5 1.0 183.7 1.0
64 16384 259.2 138.2 1.4 280.0 149.4 1.4 279.7 1.4
256 4096 352.2 140.9 1.4 352.8 141.1 1.5 352.0 1.5
512 2048 413.3 146.9 1.8 411.8 146.4 1.8 412.2 1.8
Errors are supposed to be of order of 1 (ULPs).
Also note that it must be used on Visual Studio 2008 with Service pack 1 for now.I tried to install "Service Pack 1" but it says Service Pack 1 is already included in "VS2008 Professional", operation aborted.
Sorry, for not being more precise. I'm interested in a general application that can detect the cpu optimizations. The lunatics installer was just an example.if you want to see which instruction set(SSE SSE2 SSE3 SSSE3 etc) your cpu support you can use Everest Ultimate (http://www.lavalys.com/)
I saw that some programs require Intel MKL to 'enhance' the computing capabilities and better use the 'optimizations' inside the processor. But when I saw this http://www.agner.org/optimize/blog/read.php?i=49#121 I wondered if there were any free version which could 'enhance' the mkl on my cpu, and not cripple the performance on an amd cpu.
Differnet instructions from different SSE levels built into the microprocessors may or may not be useful for given code, and in most cases simply telling the compiler to use those instructions doesn't do a very good job (i.e. is niot optimisation!)
Jason
I saw that some programs require Intel MKL to 'enhance' the computing capabilities and better use the 'optimizations' inside the processor. But when I saw this http://www.agner.org/optimize/blog/read.php?i=49#121 I wondered if there were any free version which could 'enhance' the mkl on my cpu, and not cripple the performance on an amd cpu.
(http://www.speedtest.net/result/1010976409.png) (http://www.speedtest.net)
The channel is limited by the operator, 20 megabits on downloading. Speaks: "100 you will not receive, the physiognomy will burst" (or mug crack). ;D ;D ;D
ROFL, looks like I lead with lowest ping so far ;D ;D ;D
...
By the way, the new GTX580 is already available in Germany (479,99 Euro)
...
heinz
must think about what todo
heinz
To proof full functionality some more complex project must be compiled.Used Release_vc90.sln from CUDA3.2 SDK (it contains 93 Projects)
1.6 Mio/day are precalculated. We will see if it will become true.+ 1,669,906.80 seither, today I got it.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Gesamtguthaben 2,139,379
Durchschnittliches Guthaben 73,954.83
Google? ;)Gesamtguthaben 2,139,379
Durchschnittliches Guthaben 73,954.83
Off topic. There more I see of the german translations the more I wonder who did them. They lack language intuition into both English and German ::)
22.04.2011 18:57:32 Running as a daemon
Hi Jason,Intel Composer XE 2011 Update 4
CUDA3.2 & INTEL's Compiler
Package ID: w_ccompxe_2011.1.127
I used the patch CompilerIDEPluginUpdate.zip to avoid the rules-file error sucessful.
The rules-file error is gone. ;)
Projekt: MonteCarloMultiGPU and Projekt: MonteCarlo does not sucessful compile with Intel Compiler
ptxas hung up :'(
========== Alles neu erstellen: 91 erfolgreich, Fehler bei 2, 0 übersprungen ==========
So far, we must wait for the next update.
heinz
...driver version unknown...
I wonder if it could be specific to notebooks?
... I think it would be better the app choose the GPU, than jiggle with profiles.yeah optimus has differences when you need Cuda + Graphics (DX or OpenGL), which we didn't ever need to worry about. I am curious how Boinc tells the driver version now, as it looks to sometimes give the same proper result & sometimes weird 'unknown'. I'll look into it, as really it's likely the server needs to know the proper driver version one day, to send the right applications.
my two cents....
If mine works, then it's specific to Boinc :D (i.e. they aren't using the same mechanism, I
I'll have a proper look what the use a bit later. strange they would link with nvApi and not use that. )
Mine's 64 bit, so that's eliminated as a source. Now that I'm curious I'll have a bit of a dig in the Boinc sources & give you another source patch... LoL
I'll be around for a couple of hours to run any instrumented test apps, but I've promised to go out on a call at 13:00 UTC.I'm off to work anyway. Only a few hours tonight. Heinz has just got me curious now, so I have to find out ;)
BOINC is still trying to support Win2k, likely the most recent versions of that DLL don't. Time marches on...I'll add that the nvApi interface is 'supposed' to be forward & backward compatible. I guess it seems to work for most of us, but these freaky laptop guys just never do what they are told...
Joe
Awaiting 10 Mio from pg in the next days...Got it gt540m_10Mio_primegrid (http://www.britta-d.de/images/gtx540m/gt540m_10Mio_primegrid.jpg)
waiting.....
ordered it today, time to repair the v8 Xeon now ;Dwaiting.....
If this takes too long, or this one blows up as well, I can recommend the Corsair AX1200 next.
Last nigt the machine powered suddenly off, looks like PSU is not stable.Do the heatsinks need a clean out?
Maybe the machine was overheated, CPU2 was over 60 grd celsius. I must do something with the airflow in the machine. I got the machine up again, but with a lot of trouble, disk problems etc.
After sucessful disk and filesystem-repair, the machine runs now again... we will see how long...
Still GPU work is done now, the 8 CPU's are doing nothing.
TemperaturenIt's strange that the core temps on CPU 2 are mostly lower than CPU 1, but CPU 2 has a higher overall temperature, perhaps that heatsink needs new thermal compound, or the sensor is inaccurate.
CPU1 39 °C (102 °F)
CPU2 57 °C (135 °F)
1. CPU / 1. Kern 40 °C (104 °F)
1. CPU / 2. Kern 22 °C (72 °F)
1. CPU / 3. Kern 36 °C (97 °F)
1. CPU / 4. Kern 33 °C (91 °F)
2. CPU / 1. Kern 29 °C (84 °F)
2. CPU / 2. Kern 26 °C (79 °F)
2. CPU / 3. Kern 29 °C (84 °F)
2. CPU / 4. Kern 29 °C (84 °F)
DIMM 74 °C (165 °F)
GPU Diode 68 °C (154 °F)
Temperatur 1 36 °C (97 °F)
Temperatur 2 44 °C (111 °F)
Temperatur 3 45 °C (113 °F)
FB-DIMM1 78 °C (172 °F)
FB-DIMM2 84 °C (183 °F)
FB-DIMM3 78 °C (172 °F)
FB-DIMM4 70 °C (158 °F)
ST31000340NS 35 °C (95 °F)
I could reanimate the old P4 by insert a new PSU and changed the graphic-adapter.
My P4 2,6GHz machine died, :'(
I shut it down and now never switched it on.
If I push the "on" switch nothing happened,
maybe to replace the PSU
excerpt from v0.39 installer Readme:heinz, you'll want to try the MB7_win_x86_SSE3_OpenCL_ATi_LHD4K_r390.exe app from the MB7 r390 sanity check thread, which is especially for GPUs with Max Workgroup size 128
The ATI MB application will not work on ATI cards with workgroup size 128
(e.g. HD43xx).
HD4670 has:
CL_DEVICE_MAX_WORK_GROUP_SIZE: 128
:'( :'( :'(
why ?
I'm disappointed....
GPUZ shows: gpuz_hd4670 (http://www.britta-d.de/images/ati/gpuz_hd4670.jpg)
heinz
6th of januaryCongratulations, Heinz! Keep to climb the ladder! :D
v8-Xeon pg_number_8 (http://www.britta-d.de/images/seti/pg_number_8.jpg) ;D
top_hosts (http://ww.primegrid.com/top_hosts.php)
HD4670 AGP awaiting 1 Mio pg on 28th januaryI got it on 27 january HD4670 AGP 1Mio pg (http://www.britta-d.de/images/ati/hd4670_pg_1Mio.jpg)
Yesterday V8-Xeon server get disk error on partition where BOINC is installed. I use a own 2GB partition still for BOINC. I had to reformat it and installed BOINC again. Bad block is shown in harddisk 0\DR0.Good News
---------------------------
That can´t be true Heinz.Hi Mike,
_heinz, can you do an updated AP v6 Atom app when you have a chance please, while r555 & r557 are faster than r409 on an Atom, they are nowhere near as fast as your special Atom builds,Hi Claggy, have seen it... be patient...have a lot around my ears at the moment.
See the FFTW 3.3.1 static library development thread for benches on my Atom N450
Claggy
O.K, thanks, i didn't know if you had seen it or not,_heinz, can you do an updated AP v6 Atom app when you have a chance please, while r555 & r557 are faster than r409 on an Atom, they are nowhere near as fast as your special Atom builds,Hi Claggy, have seen it... be patient...have a lot around my ears at the moment.
See the FFTW 3.3.1 static library development thread for benches on my Atom N450
Claggy
_heinz, can you do an updated AP v6 Atom app when you have a chance please, while r555 & r557 are faster than r409 on an Atom, they are nowhere near as fast as your special Atom builds,done, available in the Beta download area
See the FFTW 3.3.1 static library development thread for benches on my Atom N450
Claggy
what I have seen its is no big difference if i suspend work on all 8 cpus or not.some new measurement for V8-Xeon
will make some measurement in the next days to document it.
Ordered now a PSU tester unit to see if PSU or board is dead.PSU tester does not show anything, if I switch the power on, the display of PSU tester goes short on, then the PSU did switch the power automatic off. The PSU is definitely dead.
Waiting...
V8-Xeon is dead again.(a little late, but it is worth to tell and show picture)
14 days ago I shut down the machine to go to holidays. As I'm back I switched the power on and V8-Xeon did not start anymore. The light and the fans are on for a second and then off. On the display of the board is nothing shown. No selfest is starting. Looks like the machine eat a next PSU..... :o
_heinz
I dont mention it, my P4 with AGP HIS4670 died also, PSU burnt out the third time.Got the machine repaired after a general cleaning and demounting. :)
_heinz
You're suffering from the Wacky Nvidia GPU Memory Bug, upgrade to Boinc 7.0.36 or 7.0.38 for the partial or full fix, But Note that since 7.0.32 and later introduce a higher internal flops value for the GPU,
22.11.2012 19:23:11 | | NVIDIA GPU 0: GeForce GT 540M (driver version 306.94, CUDA version 5.0, compute capability 2.1, 1024MB, 8381384MB available, 258 GFLOPS peak)
Hi Claggy,http://boinc.berkeley.edu/dl/ should have all versions of BOINC.
I have BOINC 7.0.28x64 the actual BOINC, where I can find BOINC 7.0.38 ?
thanks in advance
_heinz
Hi Claggy,The Boinc 7 Changelog (http://boinc.berkeley.edu/dev/forum_thread.php?id=6698) thread has the changes and links for the different versions after they become available
I have BOINC 7.0.28x64 the actual BOINC, where I can find BOINC 7.0.38 ?
thanks in advance
_heinz
I installed BOINC 7.0.38 now.You're suffering from the Wacky Nvidia GPU Memory Bug, upgrade to Boinc 7.0.36 or 7.0.38 for the partial or full fix, But Note that since 7.0.32 and later introduce a higher internal flops value for the GPU,
22.11.2012 19:23:11 | | NVIDIA GPU 0: GeForce GT 540M (driver version 306.94, CUDA version 5.0, compute capability 2.1, 1024MB, 8381384MB available, 258 GFLOPS peak)
and which puts existing GPU tasks on the verge of going Maximum Time Exceeded, you should run down your GPU tasks prior to upgrading,
Claggy
V8-Xeon is dead again.V8-Xeon is back again with a LEPA 1600W PSU (http://www.lepatek.com/eng/product_content/1/1/20/#produkte), hoping it does a little bit longer than the other 3 PSU before.
14 days ago I shut down the machine to go to holidays. As I'm back I switched the power on and V8-Xeon did not start anymore. The light and the fans are on for a second and then off. On the display of the board is nothing shown. No selfest is starting. Looks like the machine eat a next PSU..... :o
_heinz
i7_gt650m_2Mio_distrrtgen (http://www.britta-d.de/images/i7/i7_gt650m_2Mio_distrrtgen.jpg)24th febrary
last not least not so bad as I thought, we should don't forget it is still a laptop.
_heinz
i7_gt650m_2Mio_distrrtgen (http://www.britta-d.de/images/i7/i7_gt650m_2Mio_distrrtgen.jpg)24th febrary
last not least not so bad as I thought, we should don't forget it is still a laptop.
_heinz
i7_gt650m_5Mio_distrrtgen (http://www.britta-d.de/images/i7/i7_gt650m_5Mio_distrrtgen.jpg)
continous work :)
Hi corsair, distrrtgen is downloading programs automatic onto your machine, nothing todo.
I`m not surprised at all.Looks like a complete redesign of the app is necessary to use Tesla`s and Titan`s properties optimal.
It was similar with Tesla back then.
i7-3630QM_primegrid_1Mio cpu-work (http://www.britta-d.de/images/i7/i7-3630QM_primegrid_1Mio.jpg)installed 2 GTX Titan EVGA SC SLI and GTX570
time goes on, V8-Xeon is nearly 4 years old, time to give it a hardware upgrade.
_heinz
Everyone knows the answer is 42.I`m not surprised at all.Looks like a complete redesign of the app is necessary to use Tesla`s and Titan`s properties optimal.
It was similar with Tesla back then.
_heinz
I know-->42 (http://en.wikipedia.org/wiki/42_%28number%29#In_The_Hitchhiker.27s_Guide_to_the_Galaxy) ;DEveryone knows the answer is 42.I`m not surprised at all.Looks like a complete redesign of the app is necessary to use Tesla`s and Titan`s properties optimal.
It was similar with Tesla back then.
_heinz
Still a lot of potential left and a lot of stuff to explore.
Is it any optimized SETI app for intel GPU??Raistmer had an Astropulse one that was being beta tested but it had some "precision" issue and was withdrawn. Search the message thread over at Seti in the Number Crunching forum. The last post was on the 12th of June.
eg Intel HD 4000 embebed in i5