We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Multiple Precision Integer GCD Performance Analysis on Parallel Architectures.
- Authors
Tembhurne, Jitendra V.; Sathe, S. R.
- Abstract
The multiple precision Greatest Common Divisor (GCD) computation is a key component in the field of computer algebra and in many cryptography applications such as public-key cryptography; in the setup phase of RSA, and in the implementation of point operations (addition, subtraction, multiplication etc.) on Elliptic Curve Cryptography (ECC), for factorization attacks and in the statistical testing of pseudo random number generators. Multiple precision computations on the CPUs are computationally expensive, but substantial performance is achieved by using OpenMP with the GNU Multiple Precision Library (GMP) and Compute Unified Device Architecture (CUDA). In this paper, we have explored the computational power of NVIDIA Graphics Processing Units (GPUs). We have implemented the Multiple Precision Integer (MPI) operations on the GPU using C CUDA. We implemented the several GCD computation algorithms using MPIs operations. Implementation results based on the CPU and GPU indicate that a significant speedup is achieved by the GPU using CUDA and OpenMP with GMP, as compared with the single core CPU implementation.
- Subjects
GREATEST common divisor; CRYPTOGRAPHY; ELLIPTIC curve cryptography; FACTORIZATION; GRAPHICS processing units
- Publication
IAENG International Journal of Computer Science, 2015, Vol 42, Issue 4, p296
- ISSN
1819-656X
- Publication type
Article