论文信息 - In-Memory Data Rearrangement for Irregular, Data-Intensive Computing

In-Memory Data Rearrangement for Irregular, Data-Intensive Computing

The data rearrangement engine (DRE) performs in-memory data restructuring to accelerate irregular, data-intensive applications. An emulation on a field-programmable gate array shows how the DRE could improve speedup, memory bandwidth, and energy consumption on three representative benchmarks.

Maya Gokhale | G. Scott Lloyd | M. Gokhale | G. S. Lloyd

[1] ChenT.,et al. Active memory cube , 2015 .

[2] Wim Dehaene,et al. A 65 nm, 850 MHz, 256 kbit, 4.3 pJ/access, Ultra Low Leakage Power Memory Using Dynamic Cell Stability and a Dual Swing Data Link , 2012, IEEE Journal of Solid-State Circuits.

[3] Maya Gokhale,et al. Processing in Memory: The Terasys Massively Parallel PIM Array , 1995, Computer.

[4] Jing Li,et al. A case for small row buffers in non-volatile main memories , 2012, 2012 IEEE 30th International Conference on Computer Design (ICCD).

[5] Samuel Williams,et al. VIRAM1: A MediaOriented Vector Processor with Embedded DRAM , 2000 .

[6] Peter M. Kogge,et al. A low cost, multithreaded processing-in-memory system , 2004, WMPI '04.

[7] Pedro C. Diniz,et al. Data reorganization engines for the next generation of system-on-a-chip FPGAs , 2002, FPGA '02.

[8] Jeffrey T. Draper,et al. A Prototype Processing-In-Memory (PIM) Chip for the Data-Intensive Architecture (DIVA) System , 2005, J. VLSI Signal Process..

[9] Zhen Fang,et al. The Impulse Memory Controller , 2001, IEEE Trans. Computers.