{"title":"访问LUT没有CAM -改进的皮尔逊哈希减少碰撞","authors":"Chung-Ping Hung, P. S. Min","doi":"10.1109/ICON.2013.6781942","DOIUrl":null,"url":null,"abstract":"While lookup table (LUT) is widely used in virtually every kind of computing devices, to achieve low access latency with low cost is always challenging. Although Content-Addressable Memory (CAM) can be used to implement LUT, it is not very cost-effective to add such an expensive device dedicated for LUT in most cases. On the other hand, using hash functions to implement LUT, which can achieve O(1) fast lookup in most cases without special hardware, is more preferable. The performance of a hash table based LUT depends on the collision rate. In this paper, we propose a hash function based on Pearson hashing which can effectively reduce the collision rate and thus improve the performance of LUT by adding a dynamically generated permutation table. We demonstrate the LUT performance can be significantly improved from very low additional cost.","PeriodicalId":219583,"journal":{"name":"2013 19th IEEE International Conference on Networks (ICON)","volume":"29 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2013-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Access LUT without CAM - Improved Pearson hashing for collision reduction\",\"authors\":\"Chung-Ping Hung, P. S. Min\",\"doi\":\"10.1109/ICON.2013.6781942\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"While lookup table (LUT) is widely used in virtually every kind of computing devices, to achieve low access latency with low cost is always challenging. Although Content-Addressable Memory (CAM) can be used to implement LUT, it is not very cost-effective to add such an expensive device dedicated for LUT in most cases. On the other hand, using hash functions to implement LUT, which can achieve O(1) fast lookup in most cases without special hardware, is more preferable. The performance of a hash table based LUT depends on the collision rate. In this paper, we propose a hash function based on Pearson hashing which can effectively reduce the collision rate and thus improve the performance of LUT by adding a dynamically generated permutation table. We demonstrate the LUT performance can be significantly improved from very low additional cost.\",\"PeriodicalId\":219583,\"journal\":{\"name\":\"2013 19th IEEE International Conference on Networks (ICON)\",\"volume\":\"29 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2013-12-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2013 19th IEEE International Conference on Networks (ICON)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/ICON.2013.6781942\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2013 19th IEEE International Conference on Networks (ICON)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICON.2013.6781942","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Access LUT without CAM - Improved Pearson hashing for collision reduction
While lookup table (LUT) is widely used in virtually every kind of computing devices, to achieve low access latency with low cost is always challenging. Although Content-Addressable Memory (CAM) can be used to implement LUT, it is not very cost-effective to add such an expensive device dedicated for LUT in most cases. On the other hand, using hash functions to implement LUT, which can achieve O(1) fast lookup in most cases without special hardware, is more preferable. The performance of a hash table based LUT depends on the collision rate. In this paper, we propose a hash function based on Pearson hashing which can effectively reduce the collision rate and thus improve the performance of LUT by adding a dynamically generated permutation table. We demonstrate the LUT performance can be significantly improved from very low additional cost.