Mohamed Bamakhrama

Mohamed Bamakhrama started the CSE Master's program in winter 2005/06 and graduated with honours in fall 2007. He did his master's thesis at the Chair for Computer Organisation and Parallel Architectures, TUM, and is now working at ST-Ericsson in Eindhoven.

Master's Thesis:
Embedded Multiprocessor System-on-Chip for Access Newtork Processing

Multicore systems are dominating the processor market; they enable the increase in computing power of a single chip in proportion to the Moore's law-driven increase in number of transistors. A similar evolution is observed in the system-on-chip (SoC) market through the emergence of multi-processor SoC (MPSoC) designs. Nevertheless, MPSoCs introduce some challenges to the system architects concerning the ecient design of memory hierarchies and system interconnects while maintaining the low power and cost constraints. In this master thesis, I tried to address some of these challenges: namely, non-cache coherent DMA transfers in MPSoCs, low instruction cache utilization by OS codes, and factors governing the system throughput in MPSoC designs.

These issues were investigated using empirical and simulation approaches. Empirical studies are conducted on the Danube platform. Danube is a commercial MPSoC platform that is based on two 32-bit MIPS cores and developed by Inneon Technologies AG for deployment in access network processing equipments such as integrated access devices, customer premises equipments, and home gateways. Simulation-based studies were conducted on a system based on the ARM MPCore architecture. Achievements include the successful implementation and testing of novel hardware and software solutions for improving the performance of non-cache coherent DMA transfers in MPSoCs. Several techniques for reducing the instruction cache miss rate were investigated and applied. Finally, a qualitative analysis of the impact of instruction reuse, number of cores, and memory bandwidth on the system throughput in MPSoC systems was presented.

Architecture of the Danube platform.

In order to approve the maturity of the coupling tool, it must be validated by the simulation of a well defined test scenario providing quantities for comparison with approved results. This leads to the second part of this work, the preparation of a fluid simulation program for the simulation of an FSI benchmarking scenario. To qualify a fluid solver for partitioned FSI simulations, it must be able to handle dynamic geometries and to transfer data from its discretization mesh to that of the coupling supervisor. Hence, numerical results were performed that show the influence of implemented features and give some first ideas about the expected results of the FSI benchmark scenarios.

Material