Linpack Xtreme Released

[hutt132]

@Regeneration If the Residual values do not match on all the runs, does that mean there is an instability?

i7 8700K

 

[hat]

Version 0.6 brings better error detection. Linpack will now stop and beep on errors.

On another note... I found it my memory OC isn't that stable after all... the new Linpack crashed on me.

Ha! I bet that came as a bit of a surprise.

 

[Regeneration]

@Regeneration If the Residual values do not match on all the runs, does that mean there is an instability?

Normally, yes, but there are freak occurrences. Intel MKL itself does the residual checksum checking.

 

[johnspack]

Command line but still not for linux, jeez....

 

[Regeneration]

@Regeneration If the Residual values do not match on all the runs, does that mean there is an instability?

After investgating... yes, it is memory-related instability.

A new version is now available.

v0.8
- Added benchmark feature.
- Added option to specify amount of threads.
- Changed the project name.

 

[notb]

There's like literally ZERO chance that I'd consciously run a lonely "exe" file shared via an anonymous file sharing platform.
Wouldn't it be easier to put it on github or something?

 

[Regeneration]

There's like literally ZERO chance that I'd consciously run a lonely "exe" file shared via an anonymous file sharing platform.
Wouldn't it be easier to put it on github or something?

https://www.softpedia.com/get/System/Benchmarks/Linpack-Xtreme.shtml

 

[notb]

https://www.softpedia.com/get/System/Benchmarks/Linpack-Xtreme.shtml

Eee... I'll pass.
You should really consider going "public" with the software. It seems you don't have any commercial plans anyway (unless it runs crypto mining during the benchmark :-D).

 

[hutt132]

After investgating... yes, it is memory-related instability.

A new version is now available.

v0.8
- Added benchmark feature.
- Added option to specify amount of threads.
- Changed the project name.

Could a CPU overclock make the memory unstable? My memory is at stock clocks.

 

[Nuckles56]

I'm finding that it automatically closes after the last test that I run, ie if I do 25 runs, it will show the results for 24 and then when the 25th run is completed, the program automatically closes on me.

 

[Regeneration]

I'm finding that it automatically closes after the last test that I run, ie if I do 25 runs, it will show the results for 24 and then when the 25th run is completed, the program automatically closes on me.

Latest version?

Did you put files in a location that requires special permissions (Program Files)? In that case, try to run as administrator.

Or another option, the entire process crashed due to instability.

 

[Nuckles56]

Latest version?

Did you put files in a location that requires special permissions (Program Files)? In that case, try to run as administrator.

Or another option, the entire process crashed due to instability.

Latest version, it is sitting in my downloads folder so that shouldn't be an issue. I hope it isn't that, as my system is running at fully stock speeds, and I can't change voltages for CPU or memory with my board.

 

[Regeneration]

Latest version, it is sitting in my downloads folder so that shouldn't be an issue. I hope it isn't that, as my system is running at fully stock speeds, and I can't change voltages for CPU or memory with my board.

Try the latest version (v0.9)... the 32-bit runs faster (less RAM), if it still occurs, maybe you have a key stuck on the keyboard.

 

[Nuckles56]

V0.9 worked successfully, so who know exactly went wrong.

 

[AlwaysHope]

Oh all these versions & updating.... I dunno!

 

[Regeneration]

Could a CPU overclock make the memory unstable? My memory is at stock clocks.

Sorry, I wasn't paying attention, you said its on a 8700K.

On modern CPUs, it is not unusual to get different residual values.

 

[Athlonite]

Oh all these versions & updating.... I dunno!

Fair suck of the sav mate it's like having to choose between prawns or steak to put on the BBQ LOL

 

[hat]

Sorry, I wasn't paying attention, you said its on a 8700K.

On modern CPUs, it is not unusual to get different residual values.

That... sounds really odd. How can you get different values and it's stable?

 

[AlwaysHope]

Fair suck of the sav mate it's like having to choose between prawns or steak to put on the BBQ LOL

We volunteer to be beta testers here, so......

 

[freeagent]

I've relied on IBT for a few years, and its been great for me. I just started using this, and its great. Too bad some of my overclocks weren't as stable as I thought, which would explain a few things. Its very sensitive, which is good. I haven't seen anything push my cpu the way this does, it loads the system pretty hard. Good job.

 

[hat]

It's kinda unsettling how many overclocks this thing has broke...

 

[kastriot]

Here are benchmarks with 0.9a ver and i needed to put static voltage for cpu because with adaptive would go up to 1.375V and was causing thermal throttle @100C, interesting this happens only with linpack libraries

 

[Regeneration]

After days of stress testing, I finally got a stable overclock... and released a new version.

v0.9.2
- Added several optimizations for AMD CPUs.
- Improved multithreading efficiency for the benchmark.
- Fixed insufficient memory error on 32-bit systems.
- Updated CPUID HWMonitor to version 1.36.
- Some minor changes.

 

[Athlonite]

- Added several optimizations for AMD CPUs.

Performs worse than previous version 74 Gflops for 0.92 vs 76 Gflops for 0.91

 

[Regeneration]

It will perform a lot better on new CPUs with lots of threads.

v0.9.1
66.1490 GFlops

v0.9.2
84.3079 GFlops

27.5 percent faster on 6c/12t CPU.