![]() ![]() Keep in mind: The CPU you choose should have a few extra threads available for Windows™ and other programs to run on to minimize any impact on the threads that NT is using. Each generation of chips change the number of cores and threads, so verify the number of threads carefully before you buy a new PC. Newer Intel I7 and I9 chips have 2 thread per core. In the future, they will probably be increased to 2 threads per core. Intel I3 and I5 chips typically have 1 thread per core. This is not optimal during those very volatile days. Windows™ and other programs will share those threads as well. If your workspace is running 6 instruments, only 2 instrument will run on their own threads, and 4 instruments will have to share threads. This is an optimal situation, because it leaves 6 threads available for Windows™ and other programs.Īn older Intel I5 has 4 cores and 4 threads (No hyper-threading like the I7 or I9 CPUs). If your workspace has 10 instruments running, each instrument will run on its own thread. An Intel I7 has 8 cores and thus 16 threads (2 threads per core). Processing an instrument's data on 2 threads(synchronous processing) would end up processing the tick data out of order. Side note, an instrument can only be processed on 1 thread to guarantee the tick data is processed linearly/asynchronous(in the correct order). The best improvement of NT 8 is that each instrument in use will be assigned to run on one of the CPU threads, separately from the other instruments if there are enough CPU threads to do so. With T/R enabled on one or two charts, NinjaTrader can easily consume 6 - 8 GB RAM. An average workspace uses 1 - 2 GB RAM, without T/R. ![]() T/R also consumes, and thus requires a lot of memory. Low daily volume instruments such as the /RTY, not so much. Instruments, such as the /ES futures, with a high average daily volume have a tremendous impact on NinjaTrader. Processing the historical data has a high impact on the CPU. Tick Replay: T/R has a very very high impact on NinjaTrader during a workspace startup, or when reloading a chart with a lot of days of data on the chart. ![]() If you are only running low impact indicators then you will probably not experience much impact running many instruments. Combine that with monitoring 8 or more instruments in your workspaces, including workspaces loaded in the background, all adds up to a high impact on the CPU. Running several chart of the same instrument, to monitor higher time-frame trends, can have a big impact especially if the other time-frames are running order flow/volume profile type of indicators. The exceptions are the Order Flow indicators. Basically, the indicators included with NinjaTrader™ are simple math formulas that calculate very quickly and are typically calculated only when the bar closes. A 3.5 GHz or faster CPU is suggested.Įxamples of low impact indicators include moving averages, oscillators, momentum, and trend vs. Generally speaking, indicators that require the Calculate property to be set to 'On each tick', such as the types mentioned previously, are going to impact the CPU the most. With the increasing volatility and trade volume in the markets, the impact will only increase. The Bid/Ask volume type of indicators are performing calculations on each incoming tick, and confluence zone indicators are analyzing large numbers of data points. In the context of computer hardware, trading style refers to the indicators your trading style uses and the number of charts and instruments your Workspaces typically have.Įxamples of high impact indicators include order-flow, volume profile, volume footprint, any type of Bid/Ask volume analysis indicators, and fibonacci/support-resistance confluence zone type indicators. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |