# Non-Algorithmic Encryption

The concept of an Encryption Algorithm has an additional twist. It is possible to encrypt data using a processing method, that compute in a more general space than Algorithms. I have studied these system in detail 1994-1999 and collected the most important results in the report Non-Algorithmic Encryption (updated and improved report 2013). I see the concept as important, as using non-algorithmic encryption, several principal difficulties can be overcome, which are problematic for the study of algorithmic encryption.

# Encryption Included in the TRNG9803 driver software

Processing of random numbers are essentially an encryption task. We now develop a new software for the TRNG9803, and plan lifting out the encryption part. The new TRNG9803 is based upon the bit-slice DES.

# A Critical Remark on Public Key Encryption

Key distribution using public key encryption (public key infrastructure) is investigated, and several problems are listed. The conclusion is that no advantage is gained when using public key: if you want Security, you must still carry keys in a secure way.

Public Key Cryptography -- WHY IT DON'T WORK AND WHY YOU DON'T NEED IT

**ASIC-RSA**, a Flexible Solution for RSA encryption on ASICs

We have access to a technology for RSA encryption implementation on ASIC:s. The background is that the popular Montgomery multiplication don't scale very well as a function of the modulus size. Before, most implementations used a 512 bit modulus. The current recommended key sizes are 1024-4096 bits, where an implementation preferably would be able to process various key sizes on the same hardware.

Using the ASIC-RSA technology for an ASIC implementation, a solution can be built that use less ASIC area compared to a solution that use Montgomery multiplication. Less area means reduced ASIC costs. Additionally, the solution can be adjusted to take a compromise between implementation chip area and execution time into account. A slower implementation can be built smaller than a fast implementation. The solution can be configured to compute one or several RSA key sizes.

The ASIC-RSA is an engineer trick, and not an algorithm, so it cannot be implemented in software. It is only of value when you build an ASIC.

To the best of my knowledge no patent has been filed on the ASIC-RSA, so a customer may optionally (and hopefully) obtain world-wide patent protection. We are willing to sell this technology exclusively to a single customer.