Flexible architecture for processing of large numbers and method therefor
Abstract:
A method of implementing large number multiplication and exponentiation is provided upon a general purpose microprocessor. These large number multiplication and exponentiation processes being common to cryptography standards such as RSA and AES that typically employ numbers with 512-bits, 1024-bits, and 2048-bits. According to the invention the method establishes the size of the large number processes according to value stored within a control register, this control register and other registers storing data are configured according to this value and accessed as N-bit registers (i.e. as 1024-bit registers for 1024-bit encryption. Additionally, the multiplication and exponentiation processes are handled according to the size of an arithmetic primitive, which is established according to the hardware configuration upon which the process is operating. As such the invention allows for an encryption process to adjust both to the configuration of the host microprocessor and supporting hardware/firmware and dynamically according to degree of security determined from the value stored within the control register.
Information query
Patent Agency Ranking
0/0