User manual VIA KT133A BROCHURE

[. . . ] VIA Technologies and Centaur Technology make no representations or warranties with respect to the accuracy or completeness of the contents of this publication or the information contained herein, and reserves the right to make changes at any time, without notice. VIA Technologies and Centaur Technology disclaim responsibility for any consequences resulting from the use of the information included herein. Microsoft and Windows are registered trademarks of Microsoft Corporation. Other product names used in this publication are for identification purposes only and may be trademarks of their respective companies. LIFE SUPPORT POLICY VIA processor products are not authorized for use as components in life support or other medical devices or systems (hereinafter called life support devices) unless a specific written agreement pertaining to such intended use is executed between the manufacturer and an officer of VIA. 1. [. . . ] They are predictable in the sense described above (i. e. , forward prediction if the initial conditions are known). Like all software algorithms, they are reproducible: they produce the same results starting with the same input Their security relies on an assumption that adversaries do not have enough computational power to perform certain mathematical operations. As a result, their irreversibility cannot be proved, although, in practice, this is not a major problem. Rather than relying solely on a software generator to produce all of the randomness, a software generator can be used in conjunction with a source of true randomness (entropy). Instead of using the previous result as the seed for the next number to be calculated, output from the independent source of entropy can be used as the seed. Even if the entropy source is imperfect (provides less than one bit of entropy per output bit), the results in this case can be more robust than what could be generated by either the software or the entropy source alone. An example is using a hardware random number generator (see the next section) to produce the seed for a software algorithm. Although a variety of designs are used, the software algorithms are often based on cryptographic hash functions. A hash algorithm mixes up its input bits in a fashion such that its output bits should be unpredictable unless the entire input is known. Thus, a good hash algorithm effectively destroys any statistical correlations among the input bits and makes it computationally infeasible to recover the input from the output. The most commonly used cryptographic hash algorithm, the Secure Hash Algorithm (SHA17 1) is a U. S. The rate varies, but this generator typically produces about 40, 000 bits per second. True Hardware Generators By now you should be convinced that good security requires good random numbers, and while it is possible to generate pretty good statistically random numbers using software, it is difficult, slow, and subject to numerous security pitfalls. The only truly random generator is some mechanism that detects quantum behavior at the sub-atomic level. This is because randomness is inherent in the behavior of sub-atomic particles (remember quantum mechanics, Heisenberg's Uncertainty Theorem, etc. ). Examples that have been used are: VIA KT133A Security Application Note - 15 The interval between the emission of particles during radioactive decays. This technique is used at a web 18 site to generate random bits that are available on-line from the decay of Krypton-85. This source generates only 30 bytes per second and requires a cumbersome (and dangerous?) pile of hardware. While implemented differently than the resistor based approach, ultimately, the source of randomness is the same. These sources have been used in a few commercially available add on random number generator devices, none of which have achieved much visibility or use. Since they are peripheral devices such as PCI cards and serial port devices, these commercial hardware generators are expensive and cumbersome. In addition, they are all made by universities or small companies with limited marketing and distribution capability. [. . . ] And unlike random tests, where a small number of test failures is to be expected, the cryptography engine is either correct or it is not. Centaur Technology has developed a comprehensive test package incorporating many of the available data sets, involving both encryption and decryption, at all three key sizes, in all four supported operating modes, with varying numbers of input blocks, using both methods of key generation, and with special tests for every corner case we have been able to think up. We have tested a number of KT133A chips in a variety of multitasking environments, with as many as 10 tasks concurrently running version of our ACE test package. No ACE failures have been observed in production steppings of the VIA KT133A. [. . . ]