--- Quote Start ---
Have you optimized c code?
And where is the bottleneck?
If you profiled your decoder you should know dataflow path.
--- Quote End ---
Yes, I know my way around the c code and was able to optimize it. However, the thing is that the clock frequency of the hard processor system is just not enough after a certain point. For example, after optimizing the code, I have achieved 15 fps. So, since I need 30 fps, I turned my direction to finding a vhdl open source for jpeg decoding that can be run on an Altera board. Do you know anything like that? If there is not any, I will have to stick with the c code, of course. Thank you for your response.