S. Manegold, P. A. Boncz, and M. L. Kersten. What Happens During a Join? Dissecting CPU and Memory Optimization Effects. In Proceedings of the 26th VLDB, pages 339--350, Sept. 2000.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....from the buffer incurring multiple random disk I Os. Furthermore, the CPU performance and memory bandwidth have increased by 50 each year (a.k.a. Moore s law) while memory latency has stayed roughly equal. Therefore, the cost of memory access increases exponentially with respect to CPU cost [34]. In this paper, we consider the join problem for massive datasets when the similarity measure can be any metric. The dataset can be composed of spatial data as well as sequence data. We approximate each page of the dataset using a Minimum Bounding Rectangle (MBR) Given two datasets, we build a ....

S. Manegold, P.A. Boncz, and M.L. Kersten. What happens during a join? Dissecting CPU and memory optimization effects. In VLDB, pages 339350, Cairo, Egypt, September 2000.

....from the buffer incurring multiple random disk I Os. Furthermore, the CPU performance and memory bandwidth have increased by 50 each year (a.k.a. Moore s law) while memory latency has stayed roughly equal. Therefore, the cost of memory access increases exponentially with respect to CPU cost [34]. In this paper, we consider the join problem for massive datasets when the similarity measure can be any metric. The dataset can be composed of spatial data as well as sequence data. We approximate each page of the dataset using a Minimum Bounding Rectangle (MBR) Given two datasets, we build a ....

S. Manegold, P.A. Boncz, and M.L. Kersten. What happens during a join? Dissecting CPU and memory optimization effects. In VLDB, pages 339--350, Cairo, Egypt, September 2000.

No context found.

S. Manegold, P. A. Boncz, and M. L. Kersten. What Happens During a Join? Dissecting CPU and Memory Optimization Effects. In Proceedings of the 26th VLDB, Sept. 2000.

....over the last decade. Hence, memory access has become a significant cost factor not only for main memory databases which cost models need to reflect. In query execution, the memory access issue has been addressed by designing new cache conscious data structures [13, 14, 1] and algorithms [15, 8]. On the modeling side, however, nothing has been published yet considering memory access appropriately. In this article, we address the problem of how to model memory access costs of database operators. As it turns out to be quite complicated to derive proper memory access cost functions for ....

....can be organized in multiple cascading levels. 2.3 Unified Hardware Model Summarizing our previous discussion, we describe a computers memory hardware as a cascading hierarchy of levels of caches (including TLBs) Each cache level is characterized by the parameters given in Table 1. In [8], we presented a system independent C program called Calibrator to measure these parameters on any computer hardware. We point out, that these parameters also cover the cost relevant characteristics of disk accesses. Hence, viewing main memory (e.g. a database system s buffer pool) as cache for ....

[Article contains additional citation context not shown here]

S. Manegold, P. A. Boncz, and M. L. Kersten. What happens during a Join? --- Dissecting CPU and Memory Optimization Effects. In Proceedings of the International Conference on Very Large Data Bases (VLDB), pages 339--350, Cairo, Egypt, September 2000.

....growth. Hence, memory access has become a significant cost factor not only for main memory databases which cost models need to reflect. In query execution, the memory access issue has been addressed by designing new cache conscious data structures [RR99, RR00, ADHS01] and algorithms [SKN94, MBK00b]. On the modeling side, however, nothing has been published yet considering memory access appropriately. In this article, we address the problem of how to model memory access costs of database operators appropriately. As it tums out to be quite complicated to derive proper memory access cost ....

....= Bi i l. To simplify the notation, we exploit the dualism that an access to level i 1 is caused a miss on level i. Introducing the miss latency li = i 1 and the respective miss bandwidth bi = fli l, we get li = Bi bi. Each cache level is characterized by the parameters given in Table 1. 2 In [MBK00b], we presented a system independent C program to measure these parameters on any computer hardware. 3 We point out, that these parameters also cover the cost relevant characteristics of disk accesses. Hence, viewing main memory (e.g. a database system s buffer pool) as cache for I O operations, ....

[Article contains additional citation context not shown here]

S. Manegold, P. A. Boncz, and M. L. Kersten. What happens during a Join? -- Dissecting CPU and Memory Optimization Effects. In Proceedings' of the International Conference on Very Large Data Bases (VLDB), pages 339-350, Cairo, Egypt, September 2000.

No context found.

S. Manegold, P. A. Boncz, and M. L. Kersten. What Happens During a Join? Dissecting CPU and Memory Optimization Effects. In Proceedings of the 26th VLDB, pages 339--350, Sept. 2000.

No context found.

S. Manegold, P. Boncz, and M. Kersten. What happens during a join? Dissecting CPU and memory optimization effects. In Proceedings of VLDB Conference, 2000.

No context found.

S. Manegold, P. Boncz, and M. Kersten. What happens during a join? Dissecting CPU and memory optimization effects. In Proceedings of VLDB Conference, 2000.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC