DSPs, BRAMs, and a Pinch of Logic: Extended Recipes for AES on FPGAs

作者：Drimer Saar^*; Gueneysu Tim; Paar Christof

来源：ACM Transactions on Reconfigurable Technology and Systems, 2010, 3(1): 3.

DOI：10.1145/1661438.1661441

• Algorithms
• Performance
• Security
• AES
• DSP blocks
• CMAC
• CTR
• FPGA
• Virtex-5

We present three lookup-table-based AES implementations that efficiently use the BlockRAM and DSP units embedded within Xilinx Virtex-5 FPGAs. An iterative module outputs a 32-bit AES round column every clock cycle, with a throughput of 1.67 Gbit/s when processing two 128-bit inputs. This construct is then replicated four times to provide a complete AES round per cycle with 6.7 Gbit/s throughput when processing eight input streams. This, in turn, is replicated ten times for a fully unrolled design providing over 52 Gbit/s of throughput. We also present implementations of a BRAM-based AES key-expansion, CMAC, and CTR modes of operation. Results for designs where DSPs are replaced by regular logic are also presented. The combination and arrangement of the specialized embedded functions available in the FPGA allows us to implement our designs using very few traditional user logic elements such as flip-flops and lookup tables, yet still achieve these high throughputs. HDL source code, simulation testbenches, and software tool commands to reproduce reported results for the three AES variants and CMAC mode are made publicly available. Our contribution concludes with a discussion on comparing cipher implementations in the literature, and why these comparisons can be meaningless without a common reporting methodology, or within the context of a constrained target application.

• 出版日期2010-1