Algorithms for Memory Hierarchies Lecture 5
نویسندگان
چکیده
Let’s suppose we have a set of one-dimensional ranges S = {s1, s2, ..., sN}, with each range in the form (x1, x2) and a set of queries Q = {q1, q2, ..., qN}. Batched Stabbing Query Problem (BSQP) asks to report for each element qi ∈ S, which ranges sj ∈ S contain qi. Another way to look at the problem is to view the set S as horizontal segments and Q as infinite vertical lines (as depicted in Figure 1). Then the BSQP asks to report for each vertical line qi, the segments that intersect qi.
منابع مشابه
Algorithms for Memory Hierarchies
The combination of parallelism and cache obliviousness is an ongoing topic of research, in this lecture we will only learn to know a few basics. In the past lectures we looked at parallel algorithms in which the memory and block size were known while in this lecture we will look at parallel algorithms in which the memory/cache size are unknown. This has many advantages like that the algorithms ...
متن کاملEATCS Bulletin, Number 83, June 2004, viii+226 pp
The existence of memory hierarchies are one of the most fundamental differences between the “classical” von Neumann model of computation and modern hardware. Therefore algorithms that take memory hierarchies into account have been intensively studied in recent years. This paper surveys basic algorithmic techniques for taking memory hierarchies into account and gives pointers to more detailed ov...
متن کامل1 . Memory Hierarchies — Models and Lower
The purpose of this introductory chapter is twofold. On the one hand, it serves the rather prosaic purpose of introducing the basic models and notations used in the subsequent chapters. On the other hand, it explains why these simple abstract models can be used to develop better algorithms for complex real world hardware. Section 1.1 starts with a basic motivation for memory hierarchies and Sec...
متن کاملDoing More for Less -- Cache-Aware Parallel Contraction Hierarchies Preprocessing
Contraction Hierarchies is a successful speedup-technique to Dijkstra’s seminal shortest path algorithm that has a convenient trade-off between preprocessing and query times. We investigate a shared-memory parallel implementation that uses O(n + m) space for storing the graph and O(1) space for each core during preprocessing. The presented data structures and algorithms consequently exploits ca...
متن کامل