LabVIEW
Memory Usage, what kind of Format to use and which Factors are decisive ?
Solved!
@MahmoudNajamy wrote:
and do you have an application in mind, where people need high speed streaming ?
Anything RF data.
Vibration measurements.
Video goes up in speed pretty fast.
Some motor testing probably as well. There are lots of things to test on a motor.
High speed is relative. It's a term like 'real time'. It doesn't really mean anything without a proper context. 500kHz might be high speed audio recording, for RF measurements it's not that much. For factory plants, 1 Hz is often adequate. So when a plant needs 100 Hz, I've seen it called high speed.
Also note that if you are going to look into RF, the RF frequency isn't (or doesn't need to be) the sampling frequency. But that's another topic.
Search LabVIEW like a graph! 
Hi Mahmoud,
@MahmoudNajamy wrote:
it shows me in csv. data for example 3,643174 or 382,441122 is this 64 bit at a number or is it in ASCII format 10bytes or 9 ?
CSV files contain just (delimited) text, usually in ASCII encoding…
(Atleast this is the convention.)
@MahmoudNajamy wrote:
Thanks for all Answers. I have another last question.
If i write in Spreadsheets array of double presicion, it shows me in csv. data for example 3,643174 or 382,441122 is this 64 bit at a number or is it in ASCII format 10bytes or 9 ?
If you convert a float (single, double or extended) to string, you'll loose information unless you write enough significant digits. The number of decimals that is enough depends on the number of bits of the float, single, double or extended.
Significant digits isn't the number of digits behind the separator. The number of digits before the separator count too.
If you want to store with binary precision, ASCII probably isn't for you. If you store enough digits, you'll get file a lot larger then needed. A binary format might be better.
@MahmoudNajamy wrote:
If i write in Spreadsheets array of double presicion,
DBL is represented in binary and converting it to a string consisting of decimal numbers is generally lossy. There are binary values that need an infinite amount of decimal digits to be represented exactly and there are simple decimal values (e.g. 0.1) that require an infinite amount of binary digits. Formatting and scanning is also a relatively expensive operation. If you want the file values to be guaranteed identical to the last bit, store them in binary. This will also make reading and writing much cheaper.
@altenbach wrote:
There are binary values that need an infinite amount of decimal digits to be represented exactly and there are simple decimal values (e.g. 0.1) that require an infinite amount of binary digits.
But, there are no binary numbers that can't be represented in ASCI with a finite number of characters, and represent the same number when converting back to binary though.
So you can convert any double to ASCII so it converts back to the same double.
If the goal is to write values in ASCI, so the values can be read back to the same values, then that's no problem.
"Memory Usage, what kind of Format to use and which Factors are decisive ?"
DDR 4 is the current standard format and Memory size and CL are the biggest factors, in some cases the chip size can affect if the motherboard can handle it and if it's single or double sided. 🐵
