Homomorphic Encryption Accelerator for Parallel NTT and BaseConv
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The commercialization of homomorphic encryption is hindered by the significant increase in ciphertext size and the high computational complexity of operations such as NTT and BaseConv, which significantly prolongs operation times.
Innovation Solution
A homomorphic encryption operation accelerator that performs NTT and BaseConv operations in parallel using a hierarchical register file structure, reducing operation time and energy consumption through efficient data layout and parallel processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If homomorphic encryption operations are performed using conventional sequential processing, then operational correctness is maintained, but operation time becomes excessively long
Solution Approach 1:
The patent divides the homomorphic encryption operation into two independent segments: NTT operation and BaseConv operation. These segmented operations are then executed in parallel using separate processing units, thereby reducing the total operation time while maintaining correctness.
Solution Approach 2:
The patent transitions from sequential one-dimensional processing to two-dimensional parallel processing by introducing independent processing paths for NTT and BaseConv operations. This dimensional change enables simultaneous execution of previously sequential operations.
2Productivity
If NTT and BaseConv operations are performed sequentially, then computational simplicity is maintained, but total operation time increases significantly
Solution Approach 1:
The processing system is segmented into distinct NTT processing units and BaseConv processing units. Each unit is specialized for its specific operation, enabling parallel execution while keeping individual unit designs relatively simple and modular.
Solution Approach 2:
The patent designs processing units that can handle different operations within their domains. The NTT processing units and BaseConv processing units together form a universal homomorphic encryption processing system that can handle various homomorphic operations through parallel specialization.
3Use of energy by moving object
If conventional processing architecture is used, then implementation simplicity is maintained, but energy consumption increases due to sequential processing
Solution Approach 1:
The register file is segmented into multiple independent banks that can be accessed simultaneously by different processing units. This segmentation enables parallel data access for NTT and BaseConv operations, reducing energy consumption by avoiding sequential memory access while the multi-bank structure remains relatively simple to implement.
Data Source
AI summary
A method of operating a homomorphic encryption operation accelerator includes performing a number theoretic transform (NTT) operation on each of first homomorphic ciphertext and second homomorphic ciphertext, and performing a base conversion operation by adding a partial sum using a first value of the NTT operation.


