I couldn't agree any more with your first 2 comments. (re: get a financial planner and listen to what he says. and the comment from coolweb about anything over 7% being fraudulent.... LOL
- mnx
markets are decentralized...
Nasdaq, NYSE, Arca, BATS, EDGX and all others have there matching engines housed at different locations and are independent from one another. However almost all of them are interconnected to each other...
FIX is not a routing system but rather a protocol for...
shouldn't that say both Linux(Redhat), FreeBSD and Windows?
;)
sorry for the nit pick, just pointing out that FreeBSD is it's own OS and not considered Linux...
- mnx
sterling's api is kinda trashy but may be able to do the job.
I'd recommend a 3.0GHz core2duo intel processor and 4GB of ram. Nothing else really matters, just toss in a cheapo video card and a cheap hard drive... (you could always do a RAID 1 config with 2 drives for data redundancy)...
It's called RTD. ;)
On another note, I've been looking into the NXCore data feed as of late. (http://www.nanex.net/NxCore/NxCore.htm)
There are no symbol limits and it's pretty efficient from a bandwidth point of view.
- mnx
the nasdaq has such files available via ftp...
ftp://ftp.nasdaqtrader.com/symboldirectory/nasdaqlisted.txt
ftp://ftp.nasdaqtrader.com/symboldirectory/otherlisted.txt
- mnx
The fireMV card is probably not worth considering.... just read this in the specs from the posted link... "Max DVI resolution 1600x1200 per display."...
Is nyse going to disseminate an indicative closing price? Or just Symbol, Side, Shares as they do now?
it's about time, the nyse gave out semi-realtime info.... still though a lot can happen in 15s...
- mnx
Tums and mnx vs. Gnome.....
well you can take our word for it, or preform some kind of benchmark.
if it's too difficult for me to preform a test you may be able to slow down your cpu and time the difference of 2 tests... ;)
do you know what a BIOS is and how to enter it?
- mnx
even "faulty software" that is running in a loop casing 100% cpu usage, is going to run that loop faster with a faster processor...
his backtests WILL run faster with a faster cpu.
- mnx
well it depends somewhat on what have now?
I think you said 2.0GHz but is that a Intel Core2Duo?
I've got a laptop with a 2.5GHz (t9300) core2duo, we could definitely time a backtest and compare the results...
given the new information adding RAM (or getting a machine with a ton more RAM isn't going to make much of a difference...
in your screenshot, TradersStudio is only using 100MB of RAM or so...
still willing to test some stuff out for you, to see what kind of speed up you could get from...
software optimizations can usually do a lot more than getting faster hardware, and yes I agree 100% cpu usage is frequently "faulty" software..... BUT ....
in this case I do not believe that it is NOT faulty software... I think the backtesting software is cpu intensive and it is fully...