… time in several graph applications. We propose to offload the stream compaction operations
to a specialized unit, the SCU. The SCU is an efficient, compact and small footprint unit that …

… GPU extension that combines both the efficient SCU and the filtering mechanism of the IRU
to improve overall graph processing efficiency. We evaluate our proposal on top of a modern …

… to a programmable Stream Compaction Unit (SCU) hardware … the graph-based algorithm
are efficiently executed on the GPU cores. The SCU is a small unit tightly integrated in the GPU …

… We analysed the characteristics of streaming graph processing and made two main observations.
… a compacting and filtering technique to prepare data for SMs of GPU for higher GPU …

… Despite these efforts, we show that irregular graph processing … SCU [8] proposes a
programmable GPU hardware extension for graph processing that is tailored to stream compaction …

… stream compaction, and propose to offload this task to a … Stream Compaction Unit (SCU)
tailored to the requirements of this kernel. The SCU is a small unit tightly integrated in the GPU …

… the performance of streaming graph processing. To improve graph update efficiency, we …
To complement adaptive batch reordering, we propose updating graphs dynamically, based …

… In the context of largescale graph processing, programmable General-Purpose Graphics
Processing Units (… GraphPEG [10] and SCU [11], on the other hand, proposed custom hardware …

… Redzone stream compaction, the first parallel stream compaction algorithm … GPU and CPU,
if k is proportionally small (k ≪ n), Redzone outperforms existing parallel stream compaction …

… González, “Scu: a gpu stream compaction unit for graph processing,” in Proceedings of
the 46th international symposium on computer architecture, 2019, pp. 424–435. [20] S. …