Computer stability under WIN10
3 replies
ctraverso
10 years ago #114556
Hello,
recently I experienced a strong instability of my computer (under win 10 with 4gb of ram) using StrategyQuant. Last month it was ok buy in January I had several crashes due to full usage of my ram, that didn’t happened neither on win7 nor on win10 last month. I didn’t change anything and I don’t understand.
Is there a way to reduce the use of ram in order to let SQ work properly?
I also noticed that sometimes the ram consumption is stable and sometimes is increasing, even developing the same strategy on different moments. For example now after the last crash, SQ works fine and RAM consumption is stable.
If I could not solve the problem I think I’ll go back to WIn7.
Thanks in advance
Claudio
0
Tomas Matejka
10 years ago #134572
You could play with Genetic Options -> Population Size setting to get less RAM load but I would definitely ‘invest’ into some additional memory. Win 10 can also have a different RAM usage policy compared to Win 7.
0
geektrader
10 years ago #134579
@tomas: that won´t really help in reducing RAM usage. With Java this totally depends on the heap you´ve assigned, not the population size. With lower allowed RAM settings, the garbage collection will kick in earlier and the program can still work then if you assign heaps of RAM in which case Java will first fill it up without collecting garbage.
@ctraverso: just go here: https://strategyquant.com/forum/topic/2968-how-to-tune-sq-for-max-performance-and-reduce-memory-usage-by-4-to-5-times/and assign the correct amount of RAM that is REALLY free for SQ via the correct command line adjustment. SQ can work with even low amounts of RAM, just that the garbage collection of Java will then kick in a lot more times and slow down SQ heavily, but it will work. Of course, having more RAM is always the best idea for the best performance.
0
ctraverso
10 years ago #134580
Thank you very much, I’ll follow your instructions. I missed that interesting post.
0
Viewing 3 replies - 1 through 3 (of 3 total)