--- Quote Start ---
The idea is that I would have a library of compression or image processing algorithms in the SD card then as I choose an algorithm to implement, it will get loaded to the FPGA so that the FPGA (or at least a portion of it) would perform the desired algorithm. So to answer the question, the whole algorithm would be loaded essentially. In this way, do the FGPA still have to access the SD card over and over again? Or can it be just one time loading?
I understand why you say that its better to store it in a RAM first, but if it will be a one time loading to the fpga, can this be ommited for the compression algorithm?
--- Quote End ---
Yes, you can do that, provided the FPGA doesn't keep accessing the SD card for any other info after the loading is completed, unless of course it needs to perform another loading.
--- Quote Start ---
Yes it will start after the first configuration by quartus, but after that the FPGA must be reconfigured without quartus. That's why I didn't understand why you said the scan chain bitstream (compression algorithm) will happen after the FPGA is already configured. What if I change the current algorithm to a bmp compression without the use of quatrus? That's when the scan chain comes in right? The new algorithm is scanned and the controller reconfigures the FPGA.
--- Quote End ---
Absolutely right! If you want to change the current algorithm to bmp for example, you just need to scan in the new algorithm bitstream using that interface, and you don't need quartus. Actually, you can even disconnect Quartus totally from your board (plug out USB-Blaster or disconnect the cable) after you've configured your FPGA the first time.
--- Quote Start ---
This should happen without the use of quartus.
--- Quote End ---
Absolutely.
Honored Contributor
