Abstract. The paradigm shift towards multicore technologies is offering a great potential of computational power for scientific and industrial applications. It is, however, posing considerable challenges to software development. This problem is impaired by increasing heterogeneity of hardware platforms on both, processor level, and by adding dedicated accelerators. Performance gains for data-and compute-intensive applications can currently only be achieved by exploiting coarse-and fine-grained parallelism on all system levels, and improved scalability with respect to constantly increasing core counts. A key methodology is hardware-aware computing where in all production steps explicit knowledge of processor, memory, and network details is profitably utilized. In this work we provide a survey on current multicore and accelerator technologies. We outline architectural features and show, how these features are exposed to the programmer and how they can be beneficially utilized in the application-mapping process. In particular, we characterize the discrepancy to conventional parallel platforms with respect to hierarchical memory sub-systems, fine-grained parallelism on several system levels, and chip-and system-level heterogeneity. We motivate the necessity of hardware-aware computing and summarize the challenges arising from high-performance heterogeneous computing. Furthermore, we investigate the interaction of hardware and application characteristics for selected applications in numerical simulation.

No abstract

No abstract

No abstract

No abstract

Cryptography was and still is one of the most interesting fields in Computer Science-related research. Where its origin lies in military and governmental use, today cryptography is widely used in everyday life. Cryptography secures communication between smart cards and card readers. It scrambles transmissions between DECT telephones and their base stations, it even entered the living rooms through digital Pay TV channels which use cryptographic methods to make sure that only their subscribers can watch their transmissions-and only what they have payed for. Similarly, each DVD player contains cryptographic techniques which were implemented to prevent unauthorized copying and playback. Also for network infrastructures, company intranets, or the global internet with its shared resources, cryptography is essential to secure transmitted data against snooping. This is especially vital to so called virtual private networks (VPNs), where a (virtually) private network is spanned using shared network resources meaning that although a shared medium is used, the spanned network behaves like a private network in terms of data security and connectivity. Even the underlying physical network infrastructure is hidden. One of the biggest problems with modern cryptography is data throughput. Combined video/audio data as on DVDs or broadcasted via Pay TV stations easily needs-depending on the quality-2 to 10 MBit/s and more which means that networks carrying this data have to be able to provide at least the same bandwidth: A smaller bandwidth would lead to visible and audible artifacts. Such applications require network devices which are not only capable of transporting the incoming data stream but also encrypt or decrypt them without becoming a bottleneck. For this reason, a number of dedicated hardware This thesis would not have been possible without the support of many people. First of all, I would like to thank my advisor Professor Dr. Arndt Bode for the great freedom he gave me to conduct my research and the excellent research environment at LRR-TUM. I also want to express my gratitude to my co-referee Professor Dr. Eike Jessen for his valuable comments and hints. The head of the Architecture Group, Dr. Wolfgang Karl, deserves credit for supporting my work. I want to especially thank him for making it possible for me to attend certain conferences which opened up more than fruitful opportunities. Here, I would especially express my gratitude to Dr. Nevin Heintze (Agere Systems) for arranging an internship, and his colleague Dr. Dino Oliva (Agere Systems). Both of them deserve credits for their invaluable input and ongoing support. I further want to thank Ed Walters (Agere Systems) for taking the time to review my thesis so that it now will look not too awkward to native English speakers. The internship would not have been possible without Dr.Tor Jeremiasson (Texas Instruments), who made contact between between Nevin Heintze and me on MICRO'33. Lastly, I want to thank Keith Bauer for providing a comfortable home during my stay in...

No abstract

No abstract