Optomized S5 SSE3

Crunchers For More Power
Crunchers For M...
Joined: 3 Aug 05
Posts: 69
Credit: 1071273
RAC: 0

standard app / 3140,11 sec


standard app / 3140,11 sec / 52,3min / result
S5T0301 SSE3 / 2724,88 sec / 45,4min / result
S5T0304 SSE / 2964,55 sec / 49,4min / result
S5T0305 SSE / 2845,25 sec / 47,4min / result
S5T0307 SSE / 2651,69 sec / 44,2min / result

All Results are valid.

gain:
S5T0301 = 15,2%
S5T0304 = 5,9%
S5T0305 = 10,4%
S5T0307 = 18,4%

System: Dualcore Opteron 170 @ 2,6 GHz

Crunchers For More Power
Crunchers For M...
Joined: 3 Aug 05
Posts: 69
Credit: 1071273
RAC: 0

@akosf: The last patches

@akosf:

The last patches are for SSE, can this later be put into the original client?

Akos Fekete
Akos Fekete
Joined: 13 Nov 05
Posts: 561
Credit: 4527270
RAC: 0

RE: @akosf: The last

Message 39060 in response to message 39059

Quote:

@akosf:

The last patches are for SSE, can this later be put into the original client?

I don't understand the question exactly.

It will not work with newer official applications, but the patches don't have time limits or patching sequences, so you can use each patch any time on the official app.

Akos Fekete
Akos Fekete
Joined: 13 Nov 05
Posts: 561
Credit: 4527270
RAC: 0

RE: S5T0307 SSE

Message 39061 in response to message 39058

Quote:

S5T0307 SSE / 2651,69 sec / 44,2min / result

All Results are valid.

S5T0307 = 18,4%

System: Dualcore Opteron 170 @ 2,6 GHz

Thanks!

So, S5T0307 will be the next stable version...

ErichZann
ErichZann
Joined: 11 Feb 05
Posts: 120
Credit: 81582
RAC: 0

hmm does anyone know if the

hmm does anyone know if the newer SSE Versions will be better for SSE3 Systems also or should one better use the "older" 301 SSE3 version?

And Akos, do you see a chance for talking with the project leaders about your idea of minimizing the fpu -> memory things where only one bit in the result is different?
(i mean sometime, dont feel like you have to hurry, we are all happy for what you do here and understand you dont have so much time very well)

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

RE: hmm does anyone know if

Message 39063 in response to message 39062

Quote:
hmm does anyone know if the newer SSE Versions will be better for SSE3 Systems also or should one better use the "older" 301 SSE3 version?

See this post. It seems that 0307 is better as 0301. But why you don't try it on your machine?

Crunchers For More Power
Crunchers For M...
Joined: 3 Aug 05
Posts: 69
Credit: 1071273
RAC: 0

S5T0308 SSE / 2431,00 sec

S5T0308 SSE / 2431,00 sec / 40,5min / result
S5T0308 SSE / 2466,48 sec / 41,1min / result

S5T0308 is very fast, but the results seemed to be invalid :-/

Akos Fekete
Akos Fekete
Joined: 13 Nov 05
Posts: 561
Credit: 4527270
RAC: 0

RE: hmm does anyone know if

Message 39065 in response to message 39062

Quote:
hmm does anyone know if the newer SSE Versions will be better for SSE3 Systems also or should one better use the "older" 301 SSE3 version?

I will do a SSE3 version from 0307, but i would like to get some results with 0308 at first.

Quote:
And Akos, do you see a chance for talking with the project leaders about your idea of minimizing the fpu -> memory things where only one bit in the result is different?

I send a mail to Bernd in connection with this thing.

Akos Fekete
Akos Fekete
Joined: 13 Nov 05
Posts: 561
Credit: 4527270
RAC: 0

RE: S5T0308 SSE / 2431,00

Message 39066 in response to message 39064

Quote:

S5T0308 SSE / 2431,00 sec / 40,5min / result
S5T0308 SSE / 2466,48 sec / 41,1min / result

S5T0308 is very fast, but the results seemed to be invalid :-/

Thanks!

So, I'm sure the validator works with zero tolerance... :-(
Bad news... we cannot reach much bigger speed-up.

Fuzzy Duck
Fuzzy Duck
Joined: 3 Dec 05
Posts: 37
Credit: 936924
RAC: 0

RE: standard app /

Message 39067 in response to message 39058

Quote:


standard app / 3140,11 sec / 52,3min / result
S5T0301 SSE3 / 2724,88 sec / 45,4min / result
S5T0304 SSE / 2964,55 sec / 49,4min / result
S5T0305 SSE / 2845,25 sec / 47,4min / result
S5T0307 SSE / 2651,69 sec / 44,2min / result

All Results are valid.

gain:
S5T0301 = 15,2%
S5T0304 = 5,9%
S5T0305 = 10,4%
S5T0307 = 18,4%

System: Dualcore Opteron 170 @ 2,6 GHz

I can also validate S5T0301. Just switched to S5T0307. Will report back in about 8 hours. I am enjoying this:-)

Comment viewing options

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