MS Windows Beta Test App 4.24 available

Bernd Machenschalk
Bernd Machenschalk
Moderator
Administrator
Joined: 15 Oct 04
Posts: 3,906
Credit: 191,160,119
RAC: 50,443
Topic 191648

A new Beta Test App is available on our usual Beta Test Page.

It features a codebase similar to the Linux 4.17 Beta Test App, with hand-coded optimized assembler calculation from Akos (and me).

See how much this speeds up things in general.

I haven't seen a single invalid result from the Linux Apps with the same analysis code, so I'm pretty sure this App won't cause any problems. Please report invalid results ASAP.

Happy crunching!

BM

BM

Gray Handcock
Gray Handcock
Joined: 11 Mar 05
Posts: 211
Credit: 135,567
RAC: 0

MS Windows Beta Test App 4.24 available

Quote:

A new Beta Test App is available on our usual Beta Test Page.

It features a codebase similar to the Linux 4.17 Beta Test App, with hand-coded optimized assembler calculation from Akos (and me).

See howm much this speeds up things in general.

I haven't seen a single invalid result from the Linux Apps with the same analysis code, so I'm pretty sure this App won't cause any problems. Please report invalid results ASAP.

Happy crunching!

BM

Well very early days still, but assuming a flat work level, this big WU should be done in around 8 hours or so, compared to 12,5 hours or so - that's very much better (grin). Now I can get some work done here in XP as well as in Linux (Dual-boot)

Gray

Gray Handcock
Gray Handcock
Joined: 11 Mar 05
Posts: 211
Credit: 135,567
RAC: 0

RE: RE: A new Beta Test

Message 43378 in response to message 43377

Quote:
Quote:

A new Beta Test App is available on our usual Beta Test Page.

It features a codebase similar to the Linux 4.17 Beta Test App, with hand-coded optimized assembler calculation from Akos (and me).

See howm much this speeds up things in general.

I haven't seen a single invalid result from the Linux Apps with the same analysis code, so I'm pretty sure this App won't cause any problems. Please report invalid results ASAP.

Happy crunching!

BM

Well very early days still, but assuming a flat work level, this big WU should be done in around 8 hours or so, compared to 12,5 hours or so - that's very much better (grin). Now I can get some work done here in XP as well as in Linux (Dual-boot)

Gray

OK 12% after an hour, so somewhat over 8 hours is a fair guess...

Gray

MRAO
MRAO
Joined: 7 May 05
Posts: 33
Credit: 11,321,305
RAC: 19,035

First result in with 4.24

First result in with 4.24 reduces 5600-5700 second range results to 3755 seconds, a 33% improvement (awaiting validation though). Next result is longer, but looking like a 46800 second result from 4.02 is heading to be about 32000 seconds with 4.24, around the same 33% level of improvement. This will get the Search Progress figures moving when it gets released! A great job Bernd and Akos!

Bernd Machenschalk
Bernd Machenschalk
Moderator
Administrator
Joined: 15 Oct 04
Posts: 3,906
Credit: 191,160,119
RAC: 50,443

RE: [Now I can get some

Message 43380 in response to message 43377

Quote:
[Now I can get some work done here in XP as well as in Linux (Dual-boot)


Would really be courious to know how the 4.24 compares to the 4.17 on the same machine. If I read you results right, there's still a big difference. Is this consistent with wall-clock time or is there a problem with reporting the CPU-time?

BM

BM

Dave Burbank
Dave Burbank
Joined: 30 Jan 06
Posts: 275
Credit: 1,548,376
RAC: 0

RE: RE: [Now I can get

Message 43381 in response to message 43380

Quote:
Quote:
[Now I can get some work done here in XP as well as in Linux (Dual-boot)

Would really be courious to know how the 4.24 compares to the 4.17 on the same machine. If I read you results right, there's still a big difference. Is this consistent with wall-clock time or is there a problem with reporting the CPU-time?

BM

I'm running a dual-boot as well, as soon as my cache is dry I'll start crunching with 4.24 to give a comparison between 4.17 and 4.24.

There are 10^11 stars in the galaxy. That used to be a huge number. But it's only a hundred billion. It's less than the national deficit! We used to call them astronomical numbers. Now we should call them economical numbers. - Richard Feynman

ca_grufti
ca_grufti
Joined: 9 Feb 05
Posts: 53
Credit: 4,309,237
RAC: 0

The stderr out for this

The stderr out for this result has a few curious/unusual "small x" entries that I've never seen before.

Valid State is "Initial" at this point so it's nothing to be terribly concerned about from my point of view.

Dave Burbank
Dave Burbank
Joined: 30 Jan 06
Posts: 275
Credit: 1,548,376
RAC: 0

RE: The stderr out for this

Message 43383 in response to message 43382

Quote:

The stderr out for this result has a few curious/unusual "small x" entries that I've never seen before.

Valid State is "Initial" at this point so it's nothing to be terribly concerned about from my point of view.

A qoute by Bernd from the Linux 4.16 Beta test page:

Quote:


It's a bit complicted to describe what it means mathematically, but this is written when a special case is called in the program. I knew that this would occur very rarely, but I wanted to see how often it actually happens in the "real world" to see if it's worth to optimize that part of the code, too.

BM

There are 10^11 stars in the galaxy. That used to be a huge number. But it's only a hundred billion. It's less than the national deficit! We used to call them astronomical numbers. Now we should call them economical numbers. - Richard Feynman

Mikie Tim T
Mikie Tim T
Joined: 22 Jan 05
Posts: 103
Credit: 207,633,459
RAC: 0

Looks like you did a bang up

Looks like you did a bang up job on the optimization! It has made a huge difference so far on my Turion X2. It's almost 50% faster! This is going to reignite interest in the project and start getting the estimated work still done number down starting this week. Great job Bernd, and keep the code coming!

ca_grufti
ca_grufti
Joined: 9 Feb 05
Posts: 53
Credit: 4,309,237
RAC: 0

Maybe that "small x" debug

Maybe that "small x" debug line could just be a counter that gets displayed once when the application wraps up and exits. It turns the stderr out into some new form of spagetti code.

ca_grufti
ca_grufti
Joined: 9 Feb 05
Posts: 53
Credit: 4,309,237
RAC: 0

This Athlon 64 takes about

This Athlon 64 takes about 5:20 to finish the type of WU that gets assigned 176.50 credits by the server.

I have no other results from this machine to compare to because it gets so hot around here in the summer that running computers voluntarily is a very bad idea indeed.

Comment viewing options

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