Conference PaperRECAPHE: REconfigurable Polynomial Modular Computation Architectures for Unified PQC and HE SchemesWang, Antian [Purdue University Fort Wayne,Department of Electrical and Computer Engineering,Fort Wayne,IN,USA,46805]; Zhang, Kaiyuan [Tufts University,Department of Electrical and Computer Engineering,Medford,MA,USA,02155]; Parhi, Keshab K [University of Minnesota,Department of Electrical and Computer Engineering,Minneapolis,MN,USA,55455]; Lao, Yingjie [Tufts University,Department of Electrical and Computer Engineering,Medford,MA,USA,02155]Post-Quantum Cryptography (PQC) and Homomorphic Encryption (HE) are emerging security primitives that strengthen data protection against adversaries equipped with quantum computing capabilities. Although PQC and HE rely on similar underlying arithmetic operations, their hardware implementations are typically developed independently due to differences in key parameters such as polynomial length and modulus bit-width. This work presents a unified lattice-based polynomial modular accelerator that efficiently supports both PQC and HE primitives, bridging these two domains toward future secure computing architectures. The proposed design introduces highly reconfigurable modular computation units that enable low-overhead runtime configuration across the parameter ranges commonly used in PQC and HE schemes. This unified architecture eliminates the need for separate domain-specific accelerators by reusing shared computation structures and workload patterns across both cryptographic schemes.IEEE2025-10-26106744981288 to 1292979-8-3315-8745-1https://doi.org/10.1109/IEEECONF67917.2025.114436102243053; 2426299Lattice-based cryptographyHomomorphic encryptionPost quantum cryptographyFPGAReconfigurable computingNational Science Foundation