**The LINPACK Thread**

[cq842000]

Smoking awesome job man... Glad it helped...

 

[Slaymate]

73.5014 GFlops | Slaymate | i7 2600K @ 5 GHz

 

[yzmxer608]

Kind of a random question, can anyone find the post for the current AMD King score? I searched for posts by CntdwnToExtn but it didn't find anything. Searched for posts with Opteron and it only came back with the first post.

 

[Slaymate]

79.8432 GFlops | CntdwnToExtn | Opty 6134 @ 2.3GHz



Uploaded with ImageShack.us

Is this the post your looking for?

 

[yzmxer608]

Thank you sir!

 

[boomstick666]

Is this LinX without AVX?
Otherwise we all should see 100+GFlops...
There is not much reason to run IBT/LinX without AVX now days...
I'll post mine later but w/AVX so it'll be around 130GFlops for me @ 4.8Ghz 1866 Ram.

Update:
133.0045 GFlops | boomstick666 | i5 2500k @ 4.8GHz

with 1866 9-10-9-28 1T.
I use LinX/IBT with AVX instructions to decide what my 24/7 clock should be.
Since it gives me 5~10c more than any other stress test.
(This is the reason I don't run 5.0 for 24/7. I can pass Prime95 fine but I can't run Linx w/AVX... it gets way too hot.... 5.0+ is only for benching for me.)
It's a good and quick way to know my new settings give me BSOD or not.
Also GFlops score is a good way to know how my CPU/RAM are doing in terms of stability.
Even if I thought it was stable, sometimes LinX w/AVX gives me lower score than before.
Especially at higher RAM tested.
Higher the RAM you test, better the score you get.
If I'm getting lower score, then there's something wrong with my OC.
If everything is right, I always score higher with any amount of RAM tested.

While it's running....

After the test.

 

[Slaymate]

@ boomstick666

The problem I have with your score is LinX 0.6.4 has been modified and it is no longer version 0.6.4
The actual app has been modified and now it cannot be faithfully replicated by our other members. Until an actual new version of LinX is released your score will not be counted. I could take the #1 spot with my LinX AVX score but I can't verify that everyone else is using the exact same benchmark.

 

[boomstick666]

@ boomstick666

The problem I have with your score is LinX 0.6.4 has been modified and it is no longer version 0.6.4
The actual app has been modified and now it cannot be faithfully replicated by our other members. Until an actual new version of LinX is released your score will not be counted. I could take the #1 spot with my LinX AVX score but I can't verify that everyone else is using the exact same benchmark.

OK. there are same kind of threads over at overclock.net and xtremesystems.
It's just a good way to compare your score with similar or other systems.
I guess they have no problem getting everyone's scores, though.

Also this one need some kind of fair rule that everyone can follow.
Especially fixed amount RAM to be tested. So everyone can compare even people with 4GB and 8+GB of ram.
Also cpu temp and cpu-z memory tab is important to show in screenies.

http://www.overclock.net/intel-cpus/1045865-sandy-bridge-battle-gflops-2-prizes.html
http://www.xtremesystems.org/forums/showthread.php?239175-Intel-4.5GHz-amp-5GHz-LinX-Stable-Club

 

[Slaymate]

OK. there are same kind of threads over at overclock.net and xtremesystems.
It's just a good way to compare your score with similar or other systems.
I guess they have no problem getting everyone's scores, though.

We don't have a problem getting everyones scores

Also this one need some kind of fair rule that everyone can follow.
Especially fixed amount RAM to be tested. So everyone can compare even people with 4GB and 8+GB of ram.
Also cpu temp and cpu-z memory tab is important to show in screenies.

Our rules are fair. You have just mistaken the purpose of the thread. This thread is all about getting the highest score you can. Tweak your system, upgrade your system, whatever it takes, I don't care. Modify your system not the benchmark.

You must run a non-modified version of LinX. I don't care what version it is, but the version must not be modified. LinX 0.6.4 does not have AVX support. If someone will modify the interface to show it's a new version then I'm good with it.

You must select the "All" Memory option.

You must successfully complete (3) passes. If you set it to make 5 passes and it fails during the 4th pass then your good to go.

This is not a who's 100% stable thread, this is a who can push their system the hardest thread.

A fixed amount of Ram to be tested? This will not happen, its absurb. If the limit is set to 4GB then what about the user with a X58 system and 6GB in triple channel. He has to remove a stick and run in dual channel mode? That's not fair. I have a (2x4GB) system, so I have to run in single channel mode or buy a (2x2GB) kit of memory? How is this fair? Especially when your saying "they have no problem getting everyone's scores, though." How is setting limits that downgrades the performance or requires a new purchase helping to get everyones score? What does my Temperature have to do with my score? If your OK with pushing your cpu into the upper 80C range then that's your decision but I personally think it's a poor decision. If your cpu is running that hot then you need to lower your overclock or upgrade your cooling method. If you want to post your temperatures go ahead but they have no actual bearing on your score other than how hot your willing to run your cpu at. The same with your memory, if you think running your memory at a higher speed is helpful then run it faster, it's your decision, but a cpu-z memory screen isn't needed.

I will give you a little advice. The faster your memory is running, the harder the memory controller has to work. The memory controller is built into the cpu, the harder it works the hotter your cpu gets. If you can run at 4.8GHz with your memory at 1866MHz then you can probably run at 4.9-5.0GHz at 1600MHz.

Your links have nothing to do with our thread. Overclock.net is having a 2500K/2600K contest, how is this getting everyone's score? And xtremesystems is a LinX Stable Club not a highest score club, they don't even post the actual score.

 

[PP Mguire]

I normally don't do these kinds of things but I figured might as well. I love benching!! Ignore the crappy temps as I'm using my backup aluminum rad. I'll OC more later when the sun isn't keeping my house at 80f.

57.8142 GFlops | PP Mguire | i5 750 @ 4.0ghz

http://img.photobucket.com/albums/v701/pp_mguire/Desktops and benchmark/linx.jpg