New Security Threats on FPGAs: From FPGA Design Tools Perspective
Sandeep Sunkavilli, Zhiming Zhang, Qiaoyan Yu
Abstract
The growing market share of FPGAs motivates the increasing number of attackers to tamper with FPGA systems. The majority of existing research efforts on FPGA security focus on counterfeiting devices, hardware Trojans, reverse engineering hardware designs via decomposing or decrypting bitstream files, and side-channel analysis attacks. Those attacks are typically limited to the FPGA systems implemented in standalone FPGA devices. As more cloud-based FPGA providers, third-party accelerator suppliers, and open-source FPGA design tools are available for prototyping, hardware acceleration, and high-performance computing, new FPGA utilization models are gradually formed. The increasing number of entities involved in the new FPGA use model leads to the emergence of new security threats and attack surfaces. Although the security issues on FPGA systems design and piracy have been widely investigated, there is limited investigation available disclosing the security threats from the FPGA design tools perspective. This work conducts a comprehensive survey on the FPGA tool security and proposes a thorough security threat landscape for the new FPGA utilization model in the era of machine learning and cloud computing.