Operand Size Reconfiguration for Big Data Processing in Memory
Paulo C. Santos1,a, Geraldo F. Oliveira1,b, Diego G. Tomé2,d, Marco A. Z. Alves2,e, Eduardo C. Almeida2,f and Luigi Carroc
1Informatics Institute - Federal University of Rio Grande do Sul - Porto Alegre, Brazil.
apcssjunior@inf.ufrgs.br
bgfojunior@inf.ufrgs.br
ccarro@inf.ufrgs.br
2Department of Informatics - Federal University of Paraná - Curitiba, Brazil.
ddgtome@inf.ufpr.br
emazalves@inf.ufpr.br
feduardo@inf.ufpr.br
ABSTRACT
Nowadays, applications that predominantly perform lookups over large databases are becoming more popular with column-stores as the database system architecture of choice. For these applications, Hybrid Memory Cubes (HMCs) can provide bandwidth of up to 320 GB/s and represents the best choice to keep the throughput for these ever increasing databases. However, even with the high available memory bandwidth and processing power, in order to achieve the peak performance, data movements through the memory hierarchy consumes an unnecessary amount of time and energy. In order to accelerate database operations, and reduce the energy consumption of the system, this paper presents the Reconfigurable Vector Unit (RVU) that enables massive and adaptive in-memory processing, extending the native HMC instructions and also increasing its effectiveness. RVU enables the programmer to reconfigure it to perform as a large vector unit or multiple small vectors units to better adjust for the application needs during different computation phases. Due to its adaptability, RVU is capable of achieving performance increase of 27× on average and reduce the DRAM energy consumption in 29% when compared to an x86 processor with 16 cores. Compared with the state-of-theart mechanism capable of performing large vector operations with fixed size, inside the HMC, RVU performed up to 12% better in terms of performance and improve in 53% the energy consumption.
Keywords: In-memory processing, Big data, Database, Reconfigurable, Vector instructions, Hybrid memory cube.
