I have been doing some reading and it appears more so now then ever an early end unit is not the signs of an unstable system..perhaps....
If you get repeated EEU then you should look at it....
However there are known reports and issues with using widely used older commands such as forceasm and forcesse with some of these newer A64 core chips...so beware....leave advmethods on but may remove the other commands and see if when the units starts it still says "SSE Boost OK"....
Also reading a bit I found that some of the newer super protein WUs (IE 1300's) have had issues in EEU...In this case errors reported as Lincs or NAN may not be a sign of instability but actually a flaw in the protein where 2 atoms get too far apart...When these occur continue on and monitor. If you consistently have more EEU or even regularly have them, then look at reducing the OC and trying again...
At the FoldingHome Site it appears the early end units now actually state that an early end my be the actual result of the simulation....
In the end any ocer should view EEU's as potential signs of instability but should take 1 EEU with a grain of salt....
hope this helps some!!!
I am into day 2 of my 7 day FH test.....In the end I will leave it running in the background 24/7....
If you get repeated EEU then you should look at it....
However there are known reports and issues with using widely used older commands such as forceasm and forcesse with some of these newer A64 core chips...so beware....leave advmethods on but may remove the other commands and see if when the units starts it still says "SSE Boost OK"....
Also reading a bit I found that some of the newer super protein WUs (IE 1300's) have had issues in EEU...In this case errors reported as Lincs or NAN may not be a sign of instability but actually a flaw in the protein where 2 atoms get too far apart...When these occur continue on and monitor. If you consistently have more EEU or even regularly have them, then look at reducing the OC and trying again...
At the FoldingHome Site it appears the early end units now actually state that an early end my be the actual result of the simulation....
In the end any ocer should view EEU's as potential signs of instability but should take 1 EEU with a grain of salt....
hope this helps some!!!
I am into day 2 of my 7 day FH test.....In the end I will leave it running in the background 24/7....
Facebook
Twitter