05-09-2018 03:15 AM - edited 05-09-2018 03:24 AM
Hello,
I am a new user in FPGA developments using OpenCL. Could someone explain the difference between 4DDR and 1 DDR platform on Virtex FPGAs?
If I use the 1DDR configuration, does it mean that I have less memory, but more resources to implement additional logic?
What is the trade-off?
Is there a documentation clarifying the differences?
Thanks for the clarification!
FpgaBob
05-09-2018 10:38 AM
Hi FPGA bob,
Is this in context of the Dynamic 5.0 DSAs?
For the Dynamic DSAs, as you use fewer DDRs you will have less total memory to work with, and you will have less potential bandwidth. 4DDRs allow you to separate read/writes per bank, but if you don't need the extra DDRs the Dynamic DSAs allow the logic that is needed for these additional controllers to be removed allowing more space for the end users designs, and likely faster build times and performance.
These advantages are covered in the release notes: https://www.xilinx.com/support/documentation/sw_manuals/xilinx2017_4/ug1238-sdx-rnil.pdf
And in Chapter 2:
Evan
05-09-2018 10:38 AM
Hi FPGA bob,
Is this in context of the Dynamic 5.0 DSAs?
For the Dynamic DSAs, as you use fewer DDRs you will have less total memory to work with, and you will have less potential bandwidth. 4DDRs allow you to separate read/writes per bank, but if you don't need the extra DDRs the Dynamic DSAs allow the logic that is needed for these additional controllers to be removed allowing more space for the end users designs, and likely faster build times and performance.
These advantages are covered in the release notes: https://www.xilinx.com/support/documentation/sw_manuals/xilinx2017_4/ug1238-sdx-rnil.pdf
And in Chapter 2:
Evan