David Kaeli - Böcker

The more rapid rate of increase in the speed of microprocessor technology than in memory speeds has created a serious 'memory gap' for computer designers and manufacturers. High Performance Memory Systems addresses this issue and examines all aspects of improving the memory system performance of general-purpose programs. Current research highlights from both industry and academia focus on: coherence, synchronization, and allocation; power-awareness, reliability, and reconfigurability; software-based memory tuning; architecture design issues; and workload considerations. Topics and features: Both hardware and software approaches to scalability and speed disparities are consideredIntroductory chapter provides broad examination of high performance memory systemsIncludes coverage of topics from several important international conferences.Edited by leading international authorities in the field, this new work provides a survey from researchers and practitioners on advances in technology, architecture, and algorithms that address scalability needs in multiprocessors and the expanding gap between CPU/network and memory speeds.It is ideally suited to researchers and R & D professionals with interests or practice in computer engineering, computer architecture, and processor architecture.

The State of Memory Technology Over the past decade there has been rapid growth in the speed of micropro­ cessors. CPU speeds are approximately doubling every eighteen months, while main memory speed doubles about every ten years. The International Tech­ nology Roadmap for Semiconductors (ITRS) study suggests that memory will remain on its current growth path. The ITRS short-and long-term targets indicate continued scaling improvements at about the current rate by 2016. This translates to bit densities increasing at two times every two years until the introduction of 8 gigabit dynamic random access memory (DRAM) chips, after which densities will increase four times every five years. A similar growth pattern is forecast for other high-density chip areas and high-performance logic (e.g., microprocessors and application specific inte­ grated circuits (ASICs)). In the future, molecular devices, 64 gigabit DRAMs and 28 GHz clock signals are targeted. Although densities continue to grow, we still do not see significant advances that will improve memory speed. These trends have created a problem that has been labeled the Memory Wall or Memory Gap.

Until now, there were few textbooks that focused on the dynamic subject of speculative execution, a topic that is crucial to the development of high performance computer architectures. Speculative Execution in High Performance Computer Architectures describes many recent advances in speculative execution techniques. It covers cutting-edge research projects, as well as numerous commercial implementations that demonstrate the value of this latency-hiding technique.The book begins with a review of control speculation techniques that use instruction cache prefetching, branch prediction and predication, and multi-path execution. It then examines dataflow speculation techniques including data cache prefetching, address value and data value speculation, pre-computation, and coherence speculation. This textbook also explores multithreaded approaches, emphasizing profile-guided speculation, speculative microarchitectures, and compiler techniques.

This book constitutes the refereed proceedings of the Second International Conference on High Performance Embedded Architectures and Compilers, HiPEAC 2007, held in Ghent, Belgium, in January 2007. The 19 revised full papers presented together with one invited keynote paper were carefully reviewed and selected from 65 submissions. The papers are organized in topical sections.