UPGRADE YOUR BROWSER

We have detected your current browser version is not the latest one. Xilinx.com uses the latest web technologies to bring you the best online experience possible. Please upgrade to a Xilinx.com supported browser:Chrome, Firefox, Internet Explorer 11, Safari. Thank you!

cancel
Showing results for 
Search instead for 
Did you mean: 
Explorer
Explorer
4,003 Views
Registered: ‎08-30-2007

PPC performance seems lower than Microblaze

Hello.

 

I'm in the feasibility stage of a project and I was testing the computational performance of PPC and Microblaze. I 'm using a 1024 FFT written in C (fixed-point). At this point I don't really care about the time that is needed to read (beginning) and write (end) all these data. I care only about the computational time. To measure the performance I do a Modelsim simulation by using the VHDL files produced by EDK. For Microblaze I'm using the ML506 board and for the PPC the ML507. I also use the highest frequency that the tool allows (without timing errors). The problem is that PPC seems to be almost 2 times slower than Microblaze which I think its not very logical. They both use internal FPGA BRAMs for Instruction and Data memories. One thing that also seems a little bit strange to me is that when I use PPC the EDK does not let me use bus frequency grater than 133.33 MHz. Maybe this is my bottleneck but I don't have experience on PPC so I need you help.I'm alsmost sure that I do something wrong but I don't know what. If you need more info just ask.

 

Thanks in advance

 

gtze

 

p.s. I also opened a Web-Case and I'm waiting the Xilinx's answer but I thought to send it here also 

0 Kudos