S5Txxxx.dat Patched App Tests - Speed up

Carlos_Pfitzner
Carlos_Pfitzner
Joined: 27 May 06
Posts: 27
Credit: 36999
RAC: 0

RE: Done some hybrid reslut

Message 40286 in response to message 40285

Quote:

Done some hybrid reslut on my A64 3500+ at standard speed
These are approx. time since they come from different WU, but all one are from the same datafile

100% with stock app. about 31800s
40% stock + 60% S5S0003 about 31600s
20% S5S0003 + 80% S5S0007 about 28100s

so a large WU only S5S0007 should get some where between 27000-27500, have too see. Over i hour faster then stock, thanks akosf.

Only the WU with stock validated yet, the two others waiting for second computer but it should probably be no problem since it´s the two stable version I have used.

Wich patch I should apply to my Atlhon XP+ ? 3Dnow! extended

On S4 I was about 1 hour/wu -:)
On S5 I not even started to crunch -> too BIG WUs to my Dialup & disk space.
->plus prevision of 240 Hours of CPU time to finish -:(
So, I aborted them, before start

Seems that u applyed patch 3 plus patch 7 ???

May be, I need to apply all patch(s) sequentially ?
-or- I need to erase the app / download a new copy, and apply only the latest patch ?

Thanks

Click signature for global team stats

LiborA
LiborA
Joined: 8 Dec 05
Posts: 74
Credit: 337135
RAC: 0

RE: Wich patch I should

Message 40287 in response to message 40286

Quote:

Wich patch I should apply to my Atlhon XP+ ? 3Dnow! extended

On S4 I was about 1 hour/wu -:)
On S5 I not even started to crunch -> too BIG WUs to my Dialup & disk space.
->plus prevision of 240 Hours of CPU time to finish -:(
So, I aborted them, before start

Seems that u applyed patch 3 plus patch 7 ???

May be, I need to apply all patch(s) sequentially ?
-or- I need to erase the app / download a new copy, and apply only the latest patch ?

Thanks

For 3DNow! instruction is not specific optimalization yet. I'm not sure but I think Athlon XP support SSE instruction (you can see it in CPU-Z) - so you can use S5S0007 optimalization.

If you decide use path so you can only download exe file from here. It's not necessary use all patches. Each patch file replace the same part of code in application - so use only your prefered patch.

BUT if you have dial-up conection to the Internet then E@H is not optimal for you. Same WUs have >16MB data file!!!

Carlos_Pfitzner
Carlos_Pfitzner
Joined: 27 May 06
Posts: 27
Credit: 36999
RAC: 0

RE: RE: Wich patch I

Message 40288 in response to message 40287

Quote:
Quote:

Wich patch I should apply to my Atlhon XP+ ? 3Dnow! extended

On S4 I was about 1 hour/wu -:)
On S5 I not even started to crunch -> too BIG WUs to my Dialup & disk space.
->plus prevision of 240 Hours of CPU time to finish -:(
So, I aborted them, before start

Seems that u applyed patch 3 plus patch 7 ???

May be, I need to apply all patch(s) sequentially ?
-or- I need to erase the app / download a new copy, and apply only the latest patch ?

Thanks

For 3DNow! instruction is not specific optimalization yet. I'm not sure but I think Athlon XP support SSE instruction (you can see it in CPU-Z) - so you can use S5S0007 optimalization.

If you decide use path so you can only download exe file from here. It's not necessary use all patches. Each patch file replace the same part of code in application - so use only your prefered patch.

BUT if you have dial-up conection to the Internet then E@H is not optimal for you. Same WUs have >16MB data file!!!

Well, my cpu can do CMOV , SSE FP , SSE , 3Dnow! , extended 3Dnow! , MMX ,
FXSAVE/FXSTOR , PSE36 , CMPXHG8
-and- cannot do SSE MMX , CLFLUSH

As I tested with a small assembly (not CPU-Z)

However on S4 days , was the 3Dnow! optimization that allowed me
to crunch the 6 hours WU into 1 Hour ... -:)
SSE Worked too, but was not as fast as 3DNow!

*Thats the reason I asked what patch is the one that place the 3DNow engine to work!
*In other wording ... may be I was not well understood.

Thanks

Click signature for global team stats

RandyC
RandyC
Joined: 18 Jan 05
Posts: 6613
Credit: 111139797
RAC: 0

RE: However on S4 days ,

Message 40289 in response to message 40288

Quote:


However on S4 days , was the 3Dnow! optimization that allowed me
to crunch the 6 hours WU into 1 Hour ... -:)
SSE Worked too, but was not as fast as 3DNow!

*Thats the reason I asked what patch is the one that place the 3DNow engine to work!
*In other wording ... may be I was not well understood.

Thanks

Presently, Akosf is not working on a 3DNow app, due to validation restrictions. This is unfortunate for me, because 2 of my 4 systems do not support SSE, but they ALL support 3DNow. I noted the same 3DNow vs SSE speed issue as you.

If you have an SSE capable system, use S5S0007. If you have SSE3 capable, use S5S0307. See this thread for links.

[edit] Hmmm. Don't see the S5S0307 there anymore. Use S5S0007 if you have SSE or higher.

Seti Classic Final Total: 11446 WU.

LiborA
LiborA
Joined: 8 Dec 05
Posts: 74
Credit: 337135
RAC: 0

RE: Here is a little

Message 40290 in response to message 40280

Quote:

Here is a little summary:

WU - Oficial app - S5T0003 - S5T0307 - S5T0308
-----
h1_0318.0_S5R1__23088_S5R1a_1 - 4012.6 sec - not tested - not tested - not tested
h1_0081.5_S5R1__242_S5R1a_1 - 4718.7 sec - 4773.2 sec - 4120.9 sec - not tested
l1_0229.0_S5R1__2610_S5R1a_0 - 3908.6 sec - 3916.4 sec - 3411.7 sec - not tested
l1_0229.0_S5R1__2610_S5R1a_1 - 3949.9 sec - 3916.1 sec - 3404.0 sec - 3117.4 sec
-----
Requested time - 100% - cca 100% - cca 87% - cca 79%

Now I finished crunching WU l1_0229.0_S5R1__2610_S5R1a_1 by S5T0711 - time: 3325.7 sec (cca 84%)

LiborA
LiborA
Joined: 8 Dec 05
Posts: 74
Credit: 337135
RAC: 0

RE: RE: Here is a little

Message 40291 in response to message 40290

Quote:
Quote:

Here is a little summary:

WU - Oficial app - S5T0003 - S5T0307 - S5T0308
-----
h1_0318.0_S5R1__23088_S5R1a_1 - 4012.6 sec - not tested - not tested - not tested
h1_0081.5_S5R1__242_S5R1a_1 - 4718.7 sec - 4773.2 sec - 4120.9 sec - not tested
l1_0229.0_S5R1__2610_S5R1a_0 - 3908.6 sec - 3916.4 sec - 3411.7 sec - not tested
l1_0229.0_S5R1__2610_S5R1a_1 - 3949.9 sec - 3916.1 sec - 3404.0 sec - 3117.4 sec
-----
Requested time - 100% - cca 100% - cca 87% - cca 79%

Now I finished crunching WU l1_0229.0_S5R1__2610_S5R1a_1 by S5T0711 - time: 3325.7 sec (cca 84%)


Now I finished crunching WU l1_0229.0_S5R1__2610_S5R1a_1 by S5T0712 - time: 3013.2 sec (cca 76%) - wow, great work, thanks Akos

James
James
Joined: 13 Apr 06
Posts: 5
Credit: 37146
RAC: 0

710 is registering a 33-34

710 is registering a 33-34 percent improvement over the 'stock' application. Very impressive.

Waiting on validation for two WUs but another two went through..

Metod, S56RKO
Metod, S56RKO
Joined: 11 Feb 05
Posts: 135
Credit: 826540260
RAC: 83677

CPU: Pentium M 755

CPU: Pentium M 755 (2GHz)

official: CPU times 4205 +- 10 secs
S5S0007: CPU times 3900 +- 5 secs (speedup 7.8%)
S5T0711: CPU times 3915 +- 5 secs (speedup 7.4%)

slight increase of time (0.4%) is not that big cost for versatility (SSE and SSE3 supported in the same binary).

[edit]
Added official times and speedup ratios
[/edit]

Metod ...

LiborA
LiborA
Joined: 8 Dec 05
Posts: 74
Credit: 337135
RAC: 0

RE: RE: RE: Here is a

Message 40294 in response to message 40291

Quote:
Quote:
Quote:

Here is a little summary:

WU - Oficial app - S5T0003 - S5T0307 - S5T0308
-----
h1_0318.0_S5R1__23088_S5R1a_1 - 4012.6 sec - not tested - not tested - not tested
h1_0081.5_S5R1__242_S5R1a_1 - 4718.7 sec - 4773.2 sec - 4120.9 sec - not tested
l1_0229.0_S5R1__2610_S5R1a_0 - 3908.6 sec - 3916.4 sec - 3411.7 sec - not tested
l1_0229.0_S5R1__2610_S5R1a_1 - 3949.9 sec - 3916.1 sec - 3404.0 sec - 3117.4 sec
-----
Requested time - 100% - cca 100% - cca 87% - cca 79%

Now I finished crunching WU l1_0229.0_S5R1__2610_S5R1a_1 by S5T0711 - time: 3325.7 sec (cca 84%)


Now I finished crunching WU l1_0229.0_S5R1__2610_S5R1a_1 by S5T0712 - time: 3013.2 sec (cca 76%) - wow, great work, thanks Akos

Now I finished crunching WU l1_0229.0_S5R1__2610_S5R1a_1 by S5T0713 - time: 2990.1 sec (cca 76%) - but the result is not identical as from official app.

Digger
Digger
Joined: 24 Mar 05
Posts: 84
Credit: 27421
RAC: 0

Akos, I am very


Akos, I am very impressed:

I have gotten a 27% speed-up on my Intel Celeron 2.93Ghz (SSE3)

Standard App ... 45,466.84 seconds
S5T0709 ........... 33,219.00 seconds

Thanks for your efforts,

Dig

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.