LINUX Beta Test App 4.16 available

Bernd Machenschalk
Bernd Machenschalk
Moderator
Administrator
Joined: 15 Oct 04
Posts: 4,273
Credit: 245,208,476
RAC: 13,236
Topic 191620

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

It features some more hand-crafted assembler code and a slightly modified (hopefully improved) CPU detection.

I've seen very large variations regarding the speed compared to the 4.01 App depending on the CPU type, so if you report speed values here, please also report the CPU.

Happy tessting!

BM

BM

Pete Yule
Pete Yule
Joined: 4 Mar 05
Posts: 5
Credit: 175,072
RAC: 0

LINUX Beta Test App 4.16 available

Here are some times. The top three are from the new science app (4.16) whereas the rest are from the 4.01 app.

10,433.72
10,112.48
10,129.39
11,911.66
11,483.37
11,483.67
11,478.31
11,625.21

New avg = 10225.20
Old avg = 11596.44

New = 0.882 * old, ie a 11.8% speedup, so far. Nice.

I'm using a P3 coppermine, 600MHz, which supports sse. It has 256 KB cache, and afaik, sse has been successfully detected by both app versions.

pete

tullio
tullio
Joined: 22 Jan 05
Posts: 2,118
Credit: 61,407,735
RAC: 0

Speed increase of about 7% on

Speed increase of about 7% on a Pentium II Deschutes, no SSE, SuSE Linux 9.3.
Tullio

ErichZann
ErichZann
Joined: 11 Feb 05
Posts: 120
Credit: 81,582
RAC: 0

why does now the

why does now the linux-version that already was faster than the windows version get improvements but the windows version doesnt?

Bernd Machenschalk
Bernd Machenschalk
Moderator
Administrator
Joined: 15 Oct 04
Posts: 4,273
Credit: 245,208,476
RAC: 13,236

RE: why does now the

Message 43090 in response to message 43089

Quote:
why does now the linux-version that already was faster than the windows version get improvements but the windows version doesnt?


We are testing this code on Linux first (this time), but the improvements will go into all x86-based Apps.

BM

BM

ErichZann
ErichZann
Joined: 11 Feb 05
Posts: 120
Credit: 81,582
RAC: 0

RE: RE: why does now the

Message 43091 in response to message 43090

Quote:
Quote:
why does now the linux-version that already was faster than the windows version get improvements but the windows version doesnt?

We are testing this code on Linux first (this time), but the improvements will go into all x86-based Apps.

BM

ah, alright, thanks.

Stef
Stef
Joined: 8 Mar 05
Posts: 206
Credit: 110,568,193
RAC: 0

11.2% speedup on a P3T

11.2% speedup on a P3T 1.26GHz

norm
norm
Joined: 9 Feb 05
Posts: 5
Credit: 288,543
RAC: 0

Athlon XP 2700+ just finished

Athlon XP 2700+ just finished first wu with 4.16. No problems so far.

4.01 approx. 25,300 sec.
4.16 22,295 sec.

50 minutes faster.

Great work.

Norm

LiborA
LiborA
Joined: 8 Dec 05
Posts: 74
Credit: 337,135
RAC: 0

After downloading beta test

After downloading beta test application 4.16 I got two validate error results. The firt was crunched by original app. The second was crunched by test app.

I crunched about 60 results on this machine and these two results are the only with any problem.

Speed-up is about 11,5% on my Athlon64 (socket 754)

tullio
tullio
Joined: 22 Jan 05
Posts: 2,118
Credit: 61,407,735
RAC: 0

Which is the meaning of

Which is the meaning of "small x" in a result report? Does it mean a short WU?
Tullio

LiborA
LiborA
Joined: 8 Dec 05
Posts: 74
Credit: 337,135
RAC: 0

The third result have

The third result have validate error too???

EDIT: All my results (by official app even test app) have blank stderr.txt, is it correct?

Comment viewing options

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