What is more important in strategy tester for speed of backtesting in visual mode.. Single Threading Rating or CPU Mark?
this is not a critical question, but in a way it is because it helps determine the type of CPU I use backtesting..
I appreciate any kind of answer and if it goes unanswered it's okay but if anyone knows it would be great because it would
help tweak the backtesting process which takes considerable time because I use many years of historical data to do my backtesting..
Thank you good friends.. You are the best!
Here are the definitions in regards to the question:
These are the ratings for a CPU I'm considering for mobile computing because I can get a deal on it..
It varies from "depends totally on the single thread rating" to "this means practically nothing"!
In Mt4 one terminal runs the strategy tester in one single thread (but you can install Mt4 several times and run the parallel each on one kernel).
MT5 can spread the optimization nearly totally on the cloud so that the performance is determined by the power of the cpu in the cloud - this is the fastest way even though it costs a little of you money.
But you can install MetaTester and offer your cpu power for the cloud (see here) and earn a little bit of money ~50$ - 100$ a year depending on you cpu power and you can use this only locally as well.
Or you can just skip it.
It's not useless at all.
Many people waste years of time on useless backtests.
I assume only for live charts , i.e : if i plug a custom backtester to run on a live chart it will use all the available horse power if need be.
The topic is about the Strategy Tester.
Anyway that's wrong, a live chart use only 1 thread.