Floatting Point on FPGA

[0033]

Hello,

I'm using a NI PXI-7833R, with a FPGA Xilinx virtex-II, i need to do some serious complicated VHDL programming on the FPGA, so for my operations i find an easier way than to make a multiplier IP , an addition IP (etc) and use 'em one by one .

it's a XilinX floatting-point core :

http://japan.xilinx.com/support/documentation/ip_documentation/floating_point_ds335.pdf

unfortunatly this document isn't 100% helpful for the study im doing before I start programming.

My questions are :

- if anyone can say anything about that, coz anything can be helpful !

- does labview support the library for the Core ?

Thanks

[0033]

one more question, can i implement AN IP CORE more than once and make 'em all function in the same time ?

i would appreciate any help !

thanks !

[AdnanZ]

I am assuming you want to programming in LabVIEW FPGA. Are you asking whether you can implement those IPs in it and then download the program on the 7833R FPGA?

[0033]

yeah exactly !

and i've been doing some calculations, ( Im going to program a filter ) and if Im gonna pass calculations through that softcore (which is faster than the normal floating points multiplying method ) , one by one , it would be impossible to be done (since i have too many arithmetic operations to do for each received data), i would need a clk of 1Ghz if not more ..
so hence my question : theory talking, is it possible to make copies of a softcore and use it in a parallel way ?

Thanks.

[AdnanZ]

In theory, you should be able to program that IP in LabVIEW FPGA. In practice, there are still some limitations in terms of what's supported on the LabVIEW FPGA software and the number of gates and space available on the FPGA hardware. I would recommend opening up LabVIEW FPGA and having a look at the functions palette, which gives all nodes that are supported in LabVIEW FPGA. Also, have a good look at the spec sheet of the 7833R hardware.

With LabVIEW FPGA, you can build an application in software and then compile without the need of having the hardware present. The compilation window will tell you whether the program you have written can be compiled for the FPGA hardware or not.

Also, LabVIEW FPGA has the ability to simulate both using random points as well as building a custom VI which gives you the ability to simulate the data you want to be sent to the program.

LabVIEW is designed to execute parallel applications, so you should be good there. Also, FPGA is truly parallel, so all your software components will run simultaneously without any jitter.

To summarize, you will need to test most of what you want to do by yourself and LabVIEW FPGA gives you the ability to do that conveniently. I would also ask you to have a look at the following page (in case you don't find the functions you want in the Functions palette on the block diagram):

IPNet - LabVIEW FPGA Functions and Example IP

[Rick_K]

We have a new method of bringing in external IP to the LabVIEW FPGA diagram, including the floating-point operations you are used to with the Xilinx tools. Check out the IP Integration Node on NILabs IP Integration Node

Download Here:

IP Integration Node

Rick Kuhlman | LabVIEW FPGA Product Manager | National Instruments | ni.com/fpga
Check out LabVIEW FPGA IPNet for browsing, downloading, and learning about LabVIEW FPGA IP Cores