Information security is of the highest importance these days, but it is quite as essential to keep the protection mechanism safe as well. After all, what good is encryption for if anyone can view the key. The Obfuscated Crypto Engine is my attempt to provide a secure cryptographic environment. There is, of course, no intention to make it as powerful as, for example, OpenSSL (or other SSL implementations) is, and it only provides basic AES encryption/decryption options (for now), but instead, to provide a working protection for the customer’s secrets.

As of this day, you may download the very first build of the engine. It is limited and only has the AES-ECB encryption functionality, but should be enough for you to test it. The hardcoded key is: 00 01 02 03 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f

There are numerous implementations of artificial neural nets. In my opinion, however, it is much more interesting, and much more practical (in certain cases) to give it a bit more personalized approach. After all, it is always good to know what goes behind the scene.

The decision was made to begin with the type of Artificial Neural Nets known as Multilayer Perceptron as the simplest one and the easiest to start with. There is no intention to fill this document with all the painful math stuff here. You may easily find it on the Internet.

There are plans to keep up the development by adding more types of Artificial Neural Nets, as well as adding genetic algorithms for finding the optimal topology and for finding the optimal fundamental parameters, such as the learing rate and momentum.

You may download the example from the link below. This example includes a test implementation which performs learning with backpropagation algorithm and trains the network to perform binary XOR operation on a pair of bits. The source code is well commented.

The FT245R controller simplifies the communication of your FPGA core with the FT245R chip. While the provided interface is very similar to that of the chip itself, the controller takes care of all the delays and other aspects of the FT245R's protocol, as well as provides separate ports for input and output.

This tiny project was made purely out of interest. The idea was to utilize the Mikroe's MINI-32 board to access a NAND Flash memory chip. The resulting program reads the contents of a NAND Flash memory chip and saves it to a file.

The ONFI Compliant NAND Controller project consists, in fact, of two parts:

• The controller itself
• Avalon Memory Mapped wrapper

The controller provides a clean and easy to use interface to a NAND Flash. It (the controller) may be used equally well with your own FPGA core or with NIOS2 processor.