• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

Bootable Linpack Xtreme with Porteus Linux Released

Regeneration

Regeneration

NGOHQ.COM
Joined
Oct 26, 2005
Messages
3,154 (0.44/day)
Stress testing unstable PCs from within Windows always posed a risk of corruption to the OS registry and files. But now, thanks to Linpack Xtreme for Linux and the lightweight Porteus Linux, you can stress test without those worries anymore. The bootable version of Linpack Xtreme integrated on Porteus Linux is just 320MB in size to fit on CD/DVD/USB.

Porteus is a lightweight and portable implementation of the Slackware Linux operating system that can be installed on a CD/DVD, MMC/SD card, USB flash drive or hard drive. Once installed on the storage media of your choice, it can be run on almost any PC, giving you the power and freedom of Linux anywhere you go.

Linpack Xtreme is a console front-end with the latest build of Linpack (Intel Math Kernel Library Benchmarks 2018.3.011) developed and maintained by ngohq.com. Linpack is a benchmark and the most aggressive stress testing software available today. Best used to test stability of overclocked PCs. Linpack tends to crash unstable PCs in a shorter period of time compared to other stress testing applications.

Linpack solves a dense (real*8) system of linear equations (Ax=b), measures the amount of time it takes to factor and solve the system, converts that time into a performance rate, and tests the results for accuracy. The generalization is in the number of equations (N) it can solve, which is not limited to 1000. Linpack uses partial pivoting to assure the accuracy of the results.

Linpack Xtreme was created because Prime95 is no longer effective like it used to be. LinX, IntelBurnTest, OCCT use outdated Linpack binaries from 2012. Modern hardware requires modern stress testing methodology with support for the latest instructions sets.

Linpack Xtreme is available for Windows, Linux, and as a bootable media. The bootable version is considered to be the most superior as the Linux SMP kernel is a lot more sensitive to hardware instabilities than Microsoft Windows. Watch this video for a short comparison of Prime95 vs. Linpack Xtreme.

Make sure to keep an eye on the temperatures as Linpack generates excessive amount of stress like never seen before.

Instructions:
1. Burn the ISO to a CD/DVD, or extract it to a USB flash drive and run the installer (x:\boot).
2. Boot from CD/DVD/USB.
3. Edit the file 'settings' to customize the stress test parameters. Define the number of runs, set problem size and leading dimensions according to the desired amount of RAM to be used:

11026 for 1GB
15825 for 2GB
22611 for 4GB
27818 for 6GB
32209 for 8GB
35000 for 9.6GB

4. Double click on run_stress_test or run_benchmark, select open and then execute in terminal.
5. Monitor temperatures from the taskbar (top right corner) or with the included run_sensors script.

Notes:
* Linpack's output will be saved in a file named results.txt.
* You can also stress test from text mode: login as guest, type 'mc' to launch Midnight Commander, navigate to the desktop folder, edit the file 'settings' and select the desired run script.
* The password for the root account is toor.
* The i586 version is meant for old hardware, it cannot access more than 8 threads and 2.5GB of RAM.

Downloads:
Bootable Linpack Xtreme with Porteus Linux i586 | Mirror #1 | Mirror #2
Bootable Linpack Xtreme with Porteus Linux x86_64 | Mirror #1 | Mirror #2
 

Attachments

  • Screenshot-1.jpg
    Screenshot-1.jpg
    654.7 KB · Views: 1,895
  • Screenshot-2.jpg
    Screenshot-2.jpg
    174.1 KB · Views: 1,599
  • Screenshot-3.jpg
    Screenshot-3.jpg
    85.2 KB · Views: 1,546
Last edited:
When I was working on the bootable Linux version of Linpack Xtreme, I encountered something intriguing. It appears the Linux SMP kernel is a lot more sensitive to CPU overclocking then Microsoft Windows.

I use the following system for my experiments: Intel Xeon W3680 @ 4.2 GHz (145x29), ASUS P6T (vanilla), 3x 4GB G.Skill Ares F3 @ 2320 MHz (10-12-12-31 CR1), and Nvidia GeForce GTX 970.

All the power saving features were disabled (SpeedStep, C1E, C-States), high-performance power plan profile set in Windows, Turbo boost off, Hyper-threading on, LLC on and manual fixed voltage for everything.

The system successfully passed (with 0 errors) the following stress tests: 24 hours of Prime95 blend, 8 hours of RealBench, 12 hours of Linpack Xtreme, AIDA64, MemTest86, MemTest64 and HCI MemTest. In addition, I ran a light load stress test to ensure idle stability due to LLC.

Now this 100% stable system crashed within seconds after running Linpack Xtreme on Linux. Kernel panic followed by a reboot.

At first, I thought it was some kind of a compatibility issue, so I tried mocking around with the BIOS configuration, and then different Linux distributions, even previous versions, but without any luck.

Eventually, after hours of testing, I found a solution: raising the CPU voltage from 1.281v to 1.325v completely solved the problem and stabilized the system.

The CPU needed an increase of nearly 50mV to be stable on Linux.
 
  • Like
Reactions: HTC
Is something wrong? I saw in the other thread you're at version 1.1.1 now, yet the "Bootable Linpack Xtreme with Porteus Linux x86_64" link takes me to the Windows version, and if I go to the mirror I wind up with a Proteus install featuring Linpack Xtreme 1.0.0.
 
Regeneration said:
Updated today to include support for AMD Ryzen 3000 series.
Click to expand...
But the ISOs are exactly the same...?

edit: Ah, only the direct links are the old files. Should be updated or removed.
 
What are other problem sizes I can use for different memory usage?
 
Can you update this to the latest version please?
 
The bootable GUI Linux has some power saving/screen saving enabled, the screen goes dark after some minutes. This is bad, because if it freezes or throws errors you won't see it.
Is there a way to disable it? It should be disabled by default IMO.

EDIT: xset s off in terminal seems to disable it. But I think I'll stick with the text mode, it seems to work just as well or better (the cycles are a tiny bit shorter).

Thanks for this tool, by the way. Haven't used it much yet, but at least once it found an error faster than P95 with Small FFTs.
 
Last edited:
Jokii said:
The bootable GUI Linux has some power saving/screen saving enabled, the screen goes dark after some minutes. This is bad, because if it freezes or throws errors you won't see it.
Is there a way to disable it? It should be disabled by default IMO.

EDIT: xset s off in terminal seems to disable it. But I think I'll stick with the text mode, it seems to work just as well or better (the cycles are a tiny bit shorter).

Thanks for this tool, by the way. Haven't used it much yet, but at least once it found an error faster than P95 with Small FFTs.
Click to expand...
It is a bit of annoyance to move the mouse once in a while, but better than having permanent system tray burns on the monitor.
 
Do modern monitors do that though? I bet they don't, or at least not after less than a few hours.
Besides, you can always turn them off.
Personally, I would remove it if it can cause errors (and thus wasting your time).
 
Octopuss said:
Do modern monitors do that though? I bet they don't, or at least not after less than a few hours.
Besides, you can always turn them off.
Personally, I would remove it if it can cause errors (and thus wasting your time).
Click to expand...

They do also. On way smaller scale, but do.
 
Octopuss said:
Do modern monitors do that though? I bet they don't, or at least not after less than a few hours.
Besides, you can always turn them off.
Personally, I would remove it if it can cause errors (and thus wasting your time).
Click to expand...
Yeah but in a smaller scale. Happened to me 1-2 years ago on ASUS VG248QE 144Hz monitor during stress testing.
 
What's the optimal amount of RAM to use for stress testing? You wrote somewhere that one shouldn't go above 35k size or ~10GB of RAM, is that still relevant? Or is it better to use the max amount available?
I'm using the bootable Linux text only mode (not Windows or the Linux GUI), so I assume paging is not an issue.

EDIT: compared 33k and 44k quickly: the latter has a higher GFlops score and higher power usage at the wall.
 
Last edited:
Jokii said:
What's the optimal amount of RAM to use for stress testing? You wrote somewhere that one shouldn't go above 35k size or ~10GB of RAM, is that still relevant? Or is it better to use the max amount available?
I'm using the bootable Linux text only mode (not Windows or the Linux GUI), so I assume paging is not an issue.
Click to expand...
That's relevant for old architectures. The more, the merrier, unless you see some kind of negative impact.
 
Alright, thanks. Yeah I think ~44k is the sweet spot for 16GB. 45k doesn't run and 44.9k runs, but with lower GFlops.

Another detail I noticed is that being logged in as root produces slightly higher GFlopfs compared to guest. I think I ran enough tests switching back and forth to get statistically meaningful results, but I'm not quite 100% yet.
It could be that being logged as guest has some additional overhead.
 
You must log in or register to reply here.
Back
Top