32 bit seems less reliable then 64 bit
Printable View
32 bit seems less reliable then 64 bit
@Dua|ist
Could you add a feature to LinX for the competition that makes it possible to run LinX for the amount of time that will be needed to run LinX stable for a valid result.
Yes of course but that doesn't matter. With 32 bit you will have lower GFlops too so it will be harder to aim for the highest GFlops. It's just to give everyone the chance to compete.Quote:
Originally Posted by Leeghoofd
exactly what i have found. longer the pass takes the more chance of finding an error.Quote:
Originally Posted by cstkl1
Those of us with more RAM get a much heavier test, as the larger amount of memory tested, the larger the problem size, & the hotter things get.
http://ense7en.com/pics/q9650/Q9650%...s%20STABLE.JPG
Hey Dualist
Are you planning on upgrading LinX to 2.0 anytime soon? I just found out my RAM might be a problem Im having and when I read the changelog of Intel Burn Test 2.0, I thought it would be a great idea to stress my system with it, but I would really like the LinX interface :D:up:
There´s some serious rock solid stability :yepp: you have there, cstkl1, time is within 203.9 and GFlops is within 62.2, probably it was so through the whole 6hrs+ session, awesome :clap: :up:
above was with kingston hyperX t1
now with dominator GT 2k C7-7-7-20 at 1.6v
http://i287.photobucket.com/albums/l...inx16vdimm.jpg
One question
Does LinX stress the RAM enough to tell if they are 100% stable? or is only to check CPU stability?
If you use 5GB+ problem sizes so it uses ALL of your system RAM, I've found LinX to be a good test for memory stability. So far it's been spot on when I've had unstable memory settings. I still use Memtest just to double-check though.
Kougar
Linpack doesn`t stress RAM a lot.
Windows MemTest 3.8 is the best RAM<--->NB stability test program.
not for x58's i7Quote:
Originally Posted by WaterFlex
i can pass memtest for the above at 1.53v
which wont even boot into windows.
I mean Windows MemTest not Memtest 86+
well bothQuote:
Originally Posted by WaterFlex
its really unreliable
bet for x58's is using prime large fft 5.4gb
u can tell within the first 30 mins.
Is that really needed? Not that it is so hard to implement, but one could possibly just look at the status panel and then stop the test if the required time has passed. I think some sort of minimum time requirement will do, without any particular upper limit.Quote:
Originally Posted by A-Grey
IntelBurnTest and LinX are not connected in any manner, they're just based on the same Linpack engine by Intel. And I haven't found anything Linpack-related in that changelog to be honest. And there were no newer Linpack files released by Intel, so LinX is still using the latest ones.Quote:
Originally Posted by slim142
As for the memory, I'm sure that using higher problem sizes in LinX is still a good way to test memory too, and there always are such specialized tools like MemTest (both DOS and Windows-versions) which will most likely do the job even better. Or a good old Prime Blend. :)
But thanks for the info on IBT. It's nice to see the competition going. The «The original. All others are copycats.» motto sounds pretty weird now though. :rolleyes: ;)
Now that there's such feature in IBT, it would be interesting to compare the GFlops numbers...
And since we're talking about updates. Here's LinX 0.5.9. Alternate link works too.
I had plenty of time to make some thorough rearrangements in the code. Not much new to spot GUI-wise however:
- elapsed/remaining time is now synchronized to Windows time
- lots of inner bug fixes and optimizations
- cosmetic interface changes
Hopefully I didn't even break anything. :)
It isn't really necessary but it would be easier to have a one click button to run the competition test.Quote:
Originally Posted by Dua|ist
I need som help here.
I have two computers here that I have been overclocking, and testing with LinX (newest verson). One of them has an Asus Commando P965 board, Q6600, 8GB mem and the other Asus P5Q Deluxe, Q9560 and 8GB mem.
The thing is that I get very inconsistent GFlops numbers with my Q9650 rigg. I get about 33 GFlops @ 3.82 Ghz, mem @ 1066 (266 strap). That is WAY to low!
With my Q6600 rigg @ 3.2 Ghz, mem @ 800 (266 strap) I get about 43 GFlops.
What can be the problem here? I'm really confused to why this happens.
Temps on the CPU is under control, so it is not throttling.
Edit : When I ran LinX I choose custom 6500MB of memory to use, and that seems to be causing my inconsistent GFlops numbers.
When I choose to use ALL memory, the resultes seem more correct. 43 GFlops for my Q6600 @ 3.2 Ghz and 54 for my Q9650 @ 3.82 Ghz.
Is this a bug or just how Linpack works?
Linpack indeed might be glitchy sometimes. For me it can repeatedly show 17-20 GFlops on some testing instead of usual 29-30 (pretty rarely though). LinX itself doesn't produce GFlops numbers, it takes them directly from Linpack's log.Quote:
Originally Posted by Betroz
I think that maybe it is the OS that has to be blamed. That is definitely not a common scenario for Windows memory manager when a process requests such huge amounts of memory at once. Maybe reducing background load will help stabilize the numbers.
Has a version of Linpack been released as a bootable ISO either for USB stick or CD yet?
If not I guess I have reason to go searching for my USB stick again and make a nice XP x64 teststick :)
cstkl1 :worship: those corsairs are no joke.
dualist, your app is really cool. if you are interested I can host your app for free . drop me a pm if interested!
Due to Duh's generousity we now have a new hosting for LinX. :up:
From now on the main download link is this: imagepedia.net/datajay/linx.zip :)
The FileDen link remains working, it just won't be updated anymore.
Dropbox link (2nd one) remains the same.
Btw, didn't anyone notice a performance decrease (GFlops-wise) in the 0.5.9? A guy at overclockers.ru reported a pretty huge drop of ~1GFlop on x32 version. Can anyone confirm this?
My pleasure dude, thank you for helping the community :) ! unless I get over 70Gb of traffic I will leave it there for unlimited time :) . I am planning to set up a server in a virtual environment with a (heavily) shared 100mbit pipe :)Quote:
Originally Posted by Dua|ist
This program looks GREAT, as soon as my rig is back in action I will be using this to stress test.
Good work :up:
Is there a plan for a Linux version of LinX? How did you get around the 50% cpu usage on Windows. On Linux, the linpack executable only used 4 cores although it reported 8 CPUs and 8 threads.
A Linux version would be cool! I will likely write one if you don't...:D Imagine running this from a linux livecd. No more lost filesystems or chkdsk!
paul_
Thanks! If everything goes well, it's gonna be even better. :)
devsk
Answered in PM. I'd like to write Linux version too, but not sure if I'm skilled enough for that right now.
Hey this is an awesome prog!! Love it.
TWO questions-
1) is it wise to just leave "problem size" at whatever the app sets it to (which seems to change depending on amount of RAM selected) ? I assume "YES" to this.
2) I can't seem to get the "create graphs" thing to work- am I missing something? I have Everest running (see icons there in the tray, screenshow below) but the option is greyed out:
http://img526.imageshack.us/img526/8758/sshot176.png
A Linux live CD version for stability testing would be awesome!
Thanks. :)Quote:
Originally Posted by LuckMan212
1) Problem size and memory are actually almost same things, just in different measuring units, sort of. So when you change one, another is changed too. That's ok.
2) You also need to enable shared memory in Everest's settings. (External applications -> tick Enable shared memory).
After all, there's somebody who's using this feature... :cool:
Frodin
Yes, but the author is somewhat a n00b, so it'll take time. Sorry.
Sounds like the ultimate dream for o'cers! :up:Quote:
Originally Posted by Frodin
LinX 0.6.0 has been released!
Same link.
Where?Quote:
Originally Posted by stasio
First page or
http://www.benchmarkhq.ru/files/LinX.7z
Stop spoiling. :)Quote:
Originally Posted by stasio
Lol. I just wanted to make sure there are no major bugs prior to announcing it here. Sorry for this delay, guys.
All right, here it is, the 0.6.0 version.
http://img352.imageshack.us/img352/5...09631pm.th.png
LinX 0.6.0 changes:
- added setting to change Linpack's priority
- removed restriction to maximium number of threads (formerly equal to the logical CPUs number)
- tweaked (rebased) the libiomp5md.dll for a higher possible Problem size of 32-bit Linpack
- added setting to add date/time to saved files (logs & screenshots), on by default
- changed progressbar to look the same on all OSes and consume less resources
- main menu changes/simplifications
- command-line keys added (experimental), for the list of available parameters launch LinX.exe /?
- some interface changes and optimizations
- easter-egg added
Mirror link is working, main link (and 1st page) will be updated soon.
Hope you'll like it. :)
Sorry,found on:Quote:
Originally Posted by Dua|ist
http://forums.overclockers.ru
and
http://www.benchmarkhq.ru
No problems, man. It's the same version, only in Russian (can be made English by deleting the local.lng file though).
Whats the easter egg?
Done, 1st page updated, links for 0.6.0: LinX.7z or LinX.zip.
It'd be great if anyone on Vista or Win7 with 3+ GB of RAM could check the maximum Problem size for 32-bit Linpack now after that dll tweak (should be a bit higher). Just edit the Linpack32 max Problem Size setting till it starts saying in its log that there's not enough memory. Thanks in advance. :)
ChaosMinionX
Well, it's an easter-egg. It just won't be fun if I tell you that straight. ;) I'm sure someone will eventually find it. For now all I can say is that it exists and that it is completely useless but somewhat cool. But not on XP. (This can be considered a hint.)
I'm wondering if there is a reason for huge differences in gflops values between different runs.
I run a E8400@4Ghz and i get very different values between runs:
When it works like it should I get around 28.5gflops and high temps on my cores but sometimes the test runs at half speed and it does only around 15gflops and temps are 10 degrees lower.
I have decent temps so it's not thermal throttling and linpack uses 100% cpu load...
could be hitting page file thats why i select max memory and then back off by 50mb and then let it carry on testing.
Dam XP x64 here so no easter egg for me. :(
What you describe is probably my fault completely, but I'm not sure yet. I'd be grateful if you could try this version and report if you still get that strange behavior or not. Same goes to all having this issue. (Should be a lot of people :()Quote:
Originally Posted by janolle
slim142
Windows7 soon will rule the PC world. But maybe I'll do another easter-egg for XP users. Once the serious stuff gets fixed.
"Linpack has stopped unexpectingly. Press the ">>" button for details"
I might be blind, but where is the ">>" button? I cannot find it anywhere on the UI, and I don't have it on my keyboard either!
It's located near the bottom-right part of the program.Quote:
Originally Posted by jobol
Weird, I dont see it on my copy running under Windows 7 x64 B1Quote:
Originally Posted by SimpleTECH
http://img23.imageshack.us/img23/7052/linxerror.png
(the little > arrow on the lower right corner is not clickable)
Work, no problem:)Quote:
Originally Posted by jobol
http://img33.imageshack.us/img33/9179/xe4000lin.jpg
Tried a few runs and it seems to work perfectly. Thx! :clap:Quote:
Originally Posted by Dua|ist
Linpack 10.2 is out on Intel's site if you find it worthy to update to as I have no idea what the differences are between 10.1 and .2.
janolle, great news, thanks for testing!
GigaByte, I'll look into that, thanks for the news.
New in Intel MKL 10.2
This release of Intel Math Kernel Library (Intel MKL 10.2) provides
optimized multithreaded performance for the newest Intel®
processors, especially the Intel® Xeon® 5500 processor.
Performance Improvements
Performance improvements cover several key math routines including
LINPACK, Out-of-core PARDISO, BLAS, and FFT. In addition, Intel
AVX (Advanced Vector Extensions) support is included for advanced
vectorization being introduced in upcoming Intel® architecture
processors. This provides support for 256-bit vector operations, in
many cases doubling performance. These are provided earlier to test
and develop forward scaling in your applications.
C#/.Net support
Intel MKL 10.2 introduces C#/.Net support examples for calling Intel
MKL functions.
BLAS
Introduce better coverage of threading in level 1 BLAS routines, in
addition to the threading for level 2 and level 3 BLAS routines.
Download
Dua|ist, do you plan to do a native 64-bit GUI? I was trying to see if I could run it in wine in Linux and I can use 32-bit wine to run 32-bit Linx which is not really useful for me. I tried to run it with 64-bit wine and it failed because it didn't know what to do with 32-bit LinX.exe (Trying to load PE image for unsupported architecture (I386)). If the GUI becomes 64-bit native, then I don't need to write one for Linux....:D I can run it under wine.
Sadly, Delphi has neither an x64 nor a Linux compiler, so making any of these versions will be almost equally hard. But x64 should be theoretically simplier. :)Quote:
Originally Posted by devsk
Congrats for the work man!
Especially the "remaining time" function is cool :-)
Im having problems if I run it at real time priority, even at stock speeds, windows just freezes a couple of seconds after the start. :confused:
Anyone else experiencing this (running on win7 64b RC/ 4gb Ram)
How to use this?
Do I have to look at the time? Lower is better?
or
The highest GFlops peaks? Higher is better?
Thanks! :)Quote:
Originally Posted by oc_junkY
Yes, this is kind of normal behaviour. ;) Consider reading the hint over the Priority setting.Quote:
Originally Posted by monza1412
I personally wouldn't set priority to anything over above normal, as it only makes the whole OS almost freeze (or completely freeze, as in your case).
In 0.6.1 (which is currently getting its last polish) LinX will have same priority as Linpack, so at least it won't freeze and you'll be able to stop it. For now stick with lower priorities.
Well, you just test it, the more GFlops the faster is your system, on higher problem sizes/memory values you're likely to get higher GFlops and higher load and higher temps.Quote:
Originally Posted by 123456iambelial
Linpack was designed as a benchmark, thus the time per test and GFlops values. If you're just stress testing/stability testing, you may simply ignore them.
roger that. Great and very useful program by the way :clap:Quote:
Originally Posted by Dua|ist
hmm, I've got an error. It says "linpack has stopped unexpectedly, press the " " button for details." Where is that button? I can't find it. I'm sure my oc is unstable, but I can't find that button.
Bottom right-hand corner.Quote:
Originally Posted by Boogerlad
Thanks.
I am using Vista 64 Bit. Which version of Linx should I run ?
I want to test my CPU for stability purposes only. Are there any settings that I need to change ? ( such as Problem Size, Priority from TaskManager etc).
Thank you:)
That button has a long story of renaming. :) This'll be fixed, but it is the one in the right part of the statusbar, as already mentioned.Quote:
Originally Posted by Boogerlad
Run Linpack 64-bit, it is somewhat more stressful and it's default on LinX launched on a 64-bit OS. I'd say, keep priority at Normal, unless you're using your computer during testing, then you might lower it to reduce the impact on the other apps you're running.Quote:
Originally Posted by Dr.Paneas
As for Problem Size/Memory, well, larger values are better, but they also need more time to complete. Play around with them, find the ones that suit you best or just use the all mem option for maximum stress (if it's gonna work for you :rolleyes:).
I found a strange issue with last LinX.
Using Win7 x64 7100 official.
Processor i7 975 8 virtual cores, 6 Gb RAM.
Set 4 cores used in LinX and 5300 Mb of RAM.
It starts correctly, I see with task manager that occupies correctly all the memory and that the average load of the 8 cpu is 50%.
After some minutes I see that every cpu go near 0% load and I expect to have the first result on LinX.
No way, the cpu remain near 0% of load, the memory taken by LinX stays at 5300 MB and nothing happens. I waited for 10 minutes and after I shut down the program that is not blocked as it accept every command.
I tested same setup with last IBT and it worked as expected.
Will LinX be updated to the newer linpack libraries?
Does anyone have a link for the lastest linpack version?
The 6.0. has a Vista64 bug.:mad:
There should be a .6.0.2 out but I cannot find a site
Alright. First of all, let me introduce a newer version: LinX 0.6.1
Main changes:
- Linpack files updated to version 10.2
- main algorithm rewritten, might improve performance a bit
- fixed an issue with inconsistent GFlops values in some cases
- hotkeys for Start and Stop buttons added: Enter and Escape keys respectively
Will update the 1st page & main mirror later.
The new 10.2 binaries were reported to have lower GFlops numbers, but good thing is they don't require external dlls. If you're concerned about GFlops, take 32-bit and/or 64-bit folders from previous LinX releases and put them into LinX folder (don't forget to delete the newer 10.2 Linpack files). But I'm sure this didn't affect the testing efficiency in any way.
Brama,
are you talking about the 0.6.1 version? Anyway, after the load of all cores went to 0% (obviously 1st pass finished as you mentioned), did it rise again (2nd pass started)? Does this issue happen with all 8 cores/some other problem size/memory values?
This new LinX uses 2 threads now with 2nd thread being idle most of the time to further lower the impact on the running Linpack. This is (imo) a big step forward, but might introduce new bugs. :(
If anyone else is having issues, please report them.
LInx 0.6 worked fine with my Vista 64 business.. 20 to 25 pass @ maximum stressQuote:
Originally Posted by Fritz_the_germ
Soory El Greco I cannot agree. Myself and many other friends of mine had a bluesreen after the second loop using an 64 bit OS.
The version 6.1 works just fine:
http://www.abload.de/img/linx_6.134n2.png
I know 3 loops are about nothing but I will run 25 loops today.
@Dual|ist,
thanks for your link
Wait a minite...you say a bluesreen after the second loop ?? Even @ default speed???
Nop, by running my oc speed. But look at this now:
http://www.abload.de/img/linx0.6.1_20_runden7mjc.jpg
Yoh guys,
using LinX0.6.1 makes me wonder why I am having different GFlops results by using the same settings.
Normally I select "All" (memory) on Vista 64 Bit and I get about 52,xxx GFlops in hardly a minute. But sometimes when I select "All" I receive about 29,xxx GFlops in a minute and about 30 seconds )on the very same OS).
http://www.abload.de/img/linx_28.07.09wxdz.png
Why do I have so large differences using the same settings and the same OS?:confused:
To be honest I don't know why this is happening, sometimes I'm getting the very same behavior on my Win7 x64. Not sure if it happens on other OSes (32-bit ones) but all reports so far have been from 64-bit Vista/Win7. And I don't remember seeing anything like this on a 32-bit XP.Quote:
Originally Posted by Fritz_the_germ
I tried to figure it out but there just doesn't seem to be any pattern in these GFlops changes. :(
I do like this program, it has became my favorite stress testing program.
Thanks Dua|ist :up:
O.K. Thank you Dua|ist.
If you get a solution for this mysterious behavior just let us know.
I agree with xpater that LinX is a very good stress programm. If LinX were runnning through your OS without errors everything is working out fine.:D
I can confirm that this sometimes happens in Windows XP X64 too.Quote:
Originally Posted by Dua|ist
Wonderful I barely crash if I am using a unstable clock now it simply stops upon error, saves me tons of time when fine-tuning:up:Quote:
Originally Posted by Dua|ist
This is great news to hear! :) But I think we better thank guys over at Intel for the Linpack 10.2. :up:Quote:
Originally Posted by Smartidiot89
Yoh guys,
I´ve another strange behaviour of my system: Running LinX0.6.1 on W 7 I get a freeze sometimes. Is a freeze caused by to low vnb oder to low vtt?
@Fritz_the_germ
Put my :2cents: that´s to low MCH volt, usually it´s for my when it happens.
Hi Gostleader,
I increased the CPU-Vtt two steps (up to 1,31v) and everything is fine now. Even the InelBurnTest ran on "maximum" without errors. DFI-Boards act sometimes that way - a little to much vnb or vtt causes the same problems like to less. It takes you a lot of time (and nervs) to find the golden voltage out.:)
@Fritz_the_germ
Nice work. :up:
Good to know about the vnb and vtt.
Yeah, I know it takes time but it feels good when one have found that sweet spot. :)
Yoh Dua|ist,
I heard there is a LinX 0.6.2 out. The link I have does not work ( LinX 0.6.2 ).:down:
Do you have a btter one?:)
I was going to report that using memory less than 16MB such as 8 or 4MB would always result in an error at the last test, that has been fixed with 0.6.2.
http://www.benchmarkhq.ru/files/LinX.7z
Something seems to be wrong. Everytime I try to hit up the link I receive some letter crap.
Yes, LinX 0.6.2 is here:
- threading and communication with Linpack improved as an attempt to fix the varying GFlops issue
- current directory is now set to LinX' .exe directory on program start to prevent getting fake "Some Linpack files missing" messages
- screenshots of main window can now be made even with Seetings or About windows open
Download links: LinX.7z or LinX.zip
Sorry for this delay, just needed to make sure there were no serious errors in there. :)
GigaByte, it's always nice to hear that something was fixed…unintentionally. :cool: :D
P.S. 1st post updated too, links should be working fine.
I noticed my OC is not more stable with LinX 0.6.1/2, the ones with the new Linpack 10.2, I have to bump the vcore compared to 0.6.0/0.5.9
Win7 X64
BTW, guys, anyone here interested in LinX with an installer? Or the current archived version suits everyone fine?
I always preffer programs that comes with installer, i find it easier to use.
Gonna try this new version, thanks Dua|ist.
Installer will probably get more people using the program. I personally like the archive approach since I determine where the app will live. However, not everyone is the same and like I said, installers are good for the masses.Quote:
Originally Posted by Dua|ist
No, I do not prefer an installer.
Reason why is because is so easy for me to just open the file, uncompress and have it ready on my desktop.
Hey, thats about same reason why i prefer soft with installers. :) Either way, all you have to do is 3-4 clicks. However without installers you might need to delete older version, rename the folder and place it manually. I personally like to keep all my programs in Program Files folder.Quote:
Originally Posted by slim142
Lets just agree it is matter of habbit.
BTW. I tried new LinX and my prime stable OC frooze on me after 10mins. I guess i'll have to bump voltage to get this 100% stable.
Well yeah, some extra clicks but for programs this small and consntantly updated, I prefer just the folder ready-to-go.Quote:
Originally Posted by esdxx
Now, if it had some kind of auto-upgrade, then installer wouldnt be a problem, but I think thats too much to ask.
I dont like extra folders in my program files folder, make the uninstall program list bigger and add more stuff in the registry.
Its all about personal preference :up:
Big yoh to you Dua|ist!Quote:
Originally Posted by Dua|ist
The answere is f*** yeah. I would realy prefer LinX with an installer. I agree totaly with Slim142:up::clap:
I thought I said I didnt want an installer :eek::confused:Quote:
Originally Posted by Fritz_the_germ
Hey, hey, nobody worry please. :)
I think I'll keep the archives and add an installer, so that everyone can use what he likes best.
And here it goes: LinX-0.6.2.1-setup.exe
This is my very 1st installer (although with NSIS it's not that hard to create one), so feel free to say what you think, like/not like, etc.
P.S. Nevermind the slightly newer version. Nothing really important there since 0.6.2, just a couple of cosmetic fixes.
Yes please continue releasing the archive version. And thanks for taking the time to make such an awesome utility Dualist.