• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

LinX / Linpack questions

Hogan773

Senior member
Wondering if someone can help me interpret. I'm using LinX to stress test my OCs.

I have 8gb of RAM. I am running LinX at either 4096 or sometimes up to "max" which is my case is 6700mb or something. 8 threads on a 2600K. Running 20 times.

I pass without errors.

MY QUESTION is that it appears my GFlops steadily slows down over the first part of the test. First pass is appx 55 GFlops, then maybe 53, then 52, 51, 50, 50 etc. I don't have the PC in front of me right now, but I believe that it sort of settles around 50 after 5-6 passes and may bump up and down after that (50.4, 50.1, 50.3, 49.9, etc)

Why does it do this initial fall? Temps increase within the first minute so its not as if we're getting great performance only when temps are lower (because the decline happens steadily over a 7-10 minute period)

Is this normal or is it a sign of some voltage deficiency or other deficiency?

Thanks for your help. I'm pretty new to this stress testing stuff but I assume others of you know a lot about how Linpack actually works.
 
Are you using Turbo?

Could also be the ram filling up.


I am overclocking to 44x, so yes I guess I'm using Turbo (is there any way not to use Turbo on SandyBridge overclocks? - if yes, I don't know how to)

Basically it just slows down its calculating speed over the first 7-10 minutes.....how does RAM filling up affect it?
 
I don't know about the Gflops issue, but you need to run more than 20 passes to get a true test of stability. I usually go 6-8 hrs minimum, at least when I'm trying out my absolute max stable OC.
 
I don't know about the Gflops issue, but you need to run more than 20 passes to get a true test of stability. I usually go 6-8 hrs minimum, at least when I'm trying out my absolute max stable OC.


Wow 8 hrs on Linpack seems like a lot of passes. Like 250 passes or something.

8 hrs Linpack probably equals a lifetime of the web surfing and light usage that my PC routinely gets......
 
I am overclocking to 44x, so yes I guess I'm using Turbo (is there any way not to use Turbo on SandyBridge overclocks? - if yes, I don't know how to)

Basically it just slows down its calculating speed over the first 7-10 minutes.....how does RAM filling up affect it?

If the Linpack process uses up all the memory and starts to push out the operating system files into the page file, it could slow it down.

Try it again using a lower memory setting and see if the same thing happens.
 
If your able to turn off HT do so, 3 things are happening one is the number of IOs you pushing to memory, Second is the flooding of memory bandwidth (Im not saying you dont have enough) there could be a choke or stall at one point and that will scue the score and Third there is another thred already going on this and it seems LinX works better wo HT.

remember HT doenst add more CPU, It just make sure to use all it has, LinX will do that already.
 
If your able to turn off HT do so, 3 things are happening one is the number of IOs you pushing to memory, Second is the flooding of memory bandwidth (Im not saying you dont have enough) there could be a choke or stall at one point and that will scue the score and Third there is another thred already going on this and it seems LinX works better wo HT.

remember HT doenst add more CPU, It just make sure to use all it has, LinX will do that already.

Interesting - thanks. I guess I don't really care since I'm just using LinX as stability test and not performance test, but I was just curious.
 
Wow 8 hrs on Linpack seems like a lot of passes. Like 250 passes or something.

8 hrs Linpack probably equals a lifetime of the web surfing and light usage that my PC routinely gets......
Does overclocking really improve performance for ' web surfing and light usage ' that your PC routinely gets ?
 
What do you think? Does everyone on this board fully utilize their PCs for productivity uses at all times such that they need to have the latest and greatest CPUs and/or OC them?

No!

I'd guess that a good majority of the people on here enjoy the hobby of futzing around with their machines and seeing how they can OC them etc. Then they go play games with them (where OCing may help a little but an expensive GPU helps much more) or run benchmarks. Oh and surf Anandtech forums and post things.

I'm just doing this for fun and so I can get some extra "free" performance with my chip. It helps when I do video encoding although again, do I really care if it takes 10 minutes or 12 minutes to do a job? No.
 
Back
Top