LinX - A simple Linpack interface

[Dua|ist]

Thank you all for your support.

slim142, sorry. I did it once again. I'll fix it I promise.

[slim142]

Waiting for the revised version ^^

[Dua|ist]

Here you go, this should work fine.

[mine]

thanx Alexander
nice program

now start the Mac OSX.10.5 version :-)

b.t.w. your name Gus(s)ev ..by chance has it something in common with the town Gussew
(district Kaliningrad former Gumbinnen ?)

Best
Mine

[Sectorz]

I not understang why I have pass 2 time 100 pass with no error and many 50 pass with no error again on the old version of linpack 1.6 and with this soft I have 1 error after 4 pass ...

[slim142]

Here you go, this should work fine.

Thank you
(Whats happening to Sectorz?)

I not understang why I have pass 2 time 100 pass with no error and many 50 pass with no error again on the old version of linpack 1.6 and with this soft I have 1 error after 4 pass ...

Hmmmmmmmmm

[Sectorz]

I have change nothing in the bios since last time, the only change on the computer is I have removed my 2 4870x2 for one 9800gx2.

Screen shot I have do for OCZ forum

[alpha0ne]

Thanks Dua|ist for the great little proggy

[Dua|ist]

b.t.w. your name Gus(s)ev ..by chance has it something in common with the town Gussew
(district Kaliningrad former Gumbinnen ?)

Well, these are just different ways of transcribing(?) the same name into English. So the town you are referring to has in Russian the same name as mine. But that's I'm afraid the only thing they have in common, as it's a rather popular name. For both towns and people.

Sectorz
Well, if there's an error - there's an error. IntelBurnTest and LinX have different ways of displaying Linpack information, but both of them are just front-ends. It's Linpack itself (well, actually your CPU) that makes an error. I'm sure that decreasing FSB a few steps or increasing the vcore a little will result in a stable pass.
Maybe it's because of different Linpack executives versions. Another graphics card might also affect stability I guess.

[Dua|ist]

Finally here's LinX 0.4.7 (Rapidshare link with full archive) with a faster new Linpack polling algorithm (which hopefully works not only on my rig ) and only one temporary file being created during testing. All those who like GFlops are invited to try the newer version.

Changelog:

- Linpack "polling" algorithm rewritten, resulting in a noticeable performance increase
- maximum allowed problem size for Linpack32 corrected
- now LinX can detect serious error even after the first test (Residual(norm) power check)

Please report about your experience with it.

P.S. The attached version will require Linpack executives, which can be downloaded separately here: http://www.intel.com/cd/software/pro...eng/266857.htm

[slim142]

Nice
Will try it asap!

[Dua|ist]

LinX 0.4.7 was so good that I decided to make it even better.
No visual changes at all, but that polling algorithm was once again rewritten and LinX should be a bit more stable now (not that it wasn't before, but... ). And now you can safely drag its window during testing.

LinX 0.4.8 : Rapidshare link (full version)
Attached version (exe only), as usual, will require Linpack executives to run.

slim142, aren't you tired of trying new versions yet? You seem to be the only one interested in this stuff.

[burebista]

Looks fine here on Vista64.

[Evilsizer]

thanks for the update!!

[dnottis]

LinX 0.4.7 was so good that I decided to make it even better.
No visual changes at all, but that polling algorithm was once again rewritten and LinX should be a bit more stable now (not that it wasn't before, but... ). And now you can safely drag its window during testing.

LinX 0.4.8 : Rapidshare link (full version)
Attached version (exe only), as usual, will require Linpack executives to run.

slim142, aren't you tired of trying new versions yet? You seem to be the only one interested in this stuff.

Err, Im sure there are many more than you know....I do

[Dua|ist]

Oh, sorry guys, I knew there must be somebody else around here. So nice to see that somebody uses LinX on this great forum.

[slim142]

LinX 0.4.7 was so good that I decided to make it even better.
No visual changes at all, but that polling algorithm was once again rewritten and LinX should be a bit more stable now (not that it wasn't before, but... ). And now you can safely drag its window during testing.

LinX 0.4.8 : Rapidshare link (full version)
Attached version (exe only), as usual, will require Linpack executives to run.

slim142, aren't you tired of trying new versions yet? You seem to be the only one interested in this stuff.

I never get tired when I like a program
When I see a new version is up, Im first to download always

I forgot to reply, with 0.4.7 I passed 100 times and it REALLY (REALLY) stressed the heck out of my system. Temps went 10c higher than with Prime95 (75c )

Will download this and leave it running later on

I JUST HOPE THIS ONE DOESNT HAVE THE BUG (You know what Im talking about...)

Err, Im sure there are many more than you know....I do

Agreed
but...

Oh, sorry guys, I knew there must be somebody else around here. So nice to see that somebody uses LinX on this great forum.

...some people are really lazy to say even "thanks"

[rge]

...some people are really lazy to say even "thanks"

That would be me

I have downloaded and tried most, I was using IBT, but this interface is easier to use and I like it stopping when an error occurs. I will use this for quickly finding stable/unstable cpu O/C's prior to running prime o/n for final.

So thanks

[celemine1Gig]

This seems to be a nice little program. Will sure come in handy when testing my overclocks soon. ATM everything's at stock, but I'll test the program the next days anyway.

Keep up the good work. It's these free little helpful tools that I love the most when it comes to Software.

[slim142]

I also like that this is "attractive" unlike IBT which, like another member said, causes eye cancer...

[OC_Newbee]

LinX 0.4.7 was so good that I decided to make it even better.
No visual changes at all, but that polling algorithm was once again rewritten and LinX should be a bit more stable now (not that it wasn't before, but... ). And now you can safely drag its window during testing.

LinX 0.4.8 : Rapidshare link (full version)
Attached version (exe only), as usual, will require Linpack executives to run.

slim142, aren't you tired of trying new versions yet? You seem to be the only one interested in this stuff.

Thanks again! this is by far the best stress proggy I've ever used.

[mine]

thanx Alex

Like the program very much

nice even for burn in new RAM and CPUs

(A.Gussev was the one who conquered first the former called German town "Gumbinnen" in Dec.1945 - now called Gussev district Kaliningrad - Hopefully we will not see such nonsense like war between Germany and Russia ever again )

best

m

[OC_Newbee]

LinX 0.4.7 was so good that I decided to make it even better.
No visual changes at all, but that polling algorithm was once again rewritten and LinX should be a bit more stable now (not that it wasn't before, but... ). And now you can safely drag its window during testing.

LinX 0.4.8 : Rapidshare link (full version)
Attached version (exe only), as usual, will require Linpack executives to run.

slim142, aren't you tired of trying new versions yet? You seem to be the only one interested in this stuff.

I think version 0.4.8 has a bug. Right when you start LinX and choose max memory (or Max which ever one) the problem size drop box went down to 0 and then immediately error out once you hit start, second try around is okay, again this is only happen when you first start LinX. I'm running this on Vista ultimate x64, not a hugh problem considering I can manually choose the highest ram option(something like 35xx mb) instead of Max, thought I let you know. As always thanks for this awesome program!

[Dua|ist]

thanx Alex

Like the program very much

nice even for burn in new RAM and CPUs

(A.Gussev was the one who conquered first the former called German town "Gumbinnen" in Dec.1945 - now called Gussev district Kaliningrad - Hopefully we will not see such nonsense like war between Germany and Russia ever again )

best

m

Thanks. I too hope there'll be no war ever again. Frieden fuer immer!

I think version 0.4.8 has a bug. Right when you start LinX and choose max memory (or Max which ever one) the problem size drop box went down to 0 and then immediately error out once you hit start, second try around is okay, again this is only happen when you first start LinX. I'm running this on Vista ultimate x64, not a hugh problem considering I can manually choose the highest ram option(something like 35xx mb) instead of Max, thought I let you know. As always thanks for this awesome program!

Thank you for the report. Just tried on WinXP, everything seems OK. I'll investigate this on Vista64 and fix it.

[celemine1Gig]

OK, today I tried the program and it seems to be a tough nut to crack, when it comes to stability testing. I mean, I know it's "only" the frontend that was designed by you Dua|ist, but it's still a really nice tool to use. Didn't have any problems with it yet.

I'll have to back down my overclock in my silent system though. While prime95 is happily running along at 3.2GHz 1.325V on this Xeon 3050, Linpack errors out after not even 2 hours. And I'm after complete stability (as far as this is possible) with this sys, as it'll be my allday multimedia rig. Anyway, thx again. Will continue to use this tool for sure.