Dynamically managed data for CPU-GPU architectures

Jablin, Thomas B.; Jablin, James A.; Prabhu, Prakash; Liu, Feng; August, David I.

doi:10.1145/2259016.2259038

Cited by 75 publications

(48 citation statements)

References 32 publications

(30 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The first category includes work on removing dynamic irregularities from applications [27] and on optimizing CPU-GPU transfers for dynamically managed data [16]. G-Streamline is a software-based runtime approach to eliminate control-flow and memory-access irregularities from GPU programs [27].…”

Section: Related Workmentioning

confidence: 99%

“…G-Streamline is a software-based runtime approach to eliminate control-flow and memory-access irregularities from GPU programs [27]. DyManD is an automatic runtime system for managing recursive data structures (like trees) on GPUs [16]. The second category includes GPU-specific optimized implementations of breadth-first search [15], [20], single-source shortest paths [18], points-to analysis [19] and n-body simulation [6].…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Data-Driven Versus Topology-driven Irregular Computations on GPUs

Nasre

Burtscher

Pingali

2013

2013 IEEE 27th International Symposium on Parallel and Distributed Processing

View full text Add to dashboard Cite

Abstract-Irregular algorithms are algorithms with complex main data structures such as directed and undirected graphs, trees, etc. A useful abstraction for many irregular algorithms is its operator formulation in which the algorithm is viewed as the iterated application of an operator to certain nodes, called active nodes, in the graph. Each operator application, called an activity, usually touches only a small part of the overall graph, so nonoverlapping activities can be performed in parallel. In topologydriven implementations, all nodes are assumed to be active so the operator is applied everywhere in the graph even if there is no work to do at some nodes. In contrast, in data-driven implementations the operator is applied only to nodes at which there might be work to do. Multicore implementations of irregular algorithms are usually data-driven because current multicores only support small numbers of threads and work-efficiency is important. Conversely, many irregular GPU implementations use a topology-driven approach because work inefficiency can be counterbalanced by the large number of GPU threads.In this paper, we study data-driven and topology-driven implementations of six important graph algorithms on GPUs. Our goal is to understand the tradeoffs between these implementations and how to optimize them. We find that data-driven versions are generally faster and scale better despite the cost of maintaining a worklist. However, topology-driven versions can be superior when certain algorithmic properties are exploited to optimize the implementation. These results led us to devise hybrid approaches that combine the two techniques and outperform both of them.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Data-Driven Versus Topology-driven Irregular Computations on GPUs

Nasre

Burtscher

Pingali

2013

2013 IEEE 27th International Symposium on Parallel and Distributed Processing

View full text Add to dashboard Cite

show abstract

“…Overall it is 3.72x (geomean) slower than manual. DyManD [9] extends CGCM with a memory optimization framework similar to ADSM. However its state machine is considerably less detailed than that of ADSM -tracking only ownership but not coherence state.…”

Section: Comparison With Cetus+openmpc Adsm and Dymandmentioning

confidence: 99%

“…Although there are several proposals that can automate CPU-GPU memory management [6,8,9,11], our analysis (Section 3) shows that all of them introduce redundant transfers. These transfers increase the execution time of programs when compared to the use of hand-tuned manual transfers.…”

Section: Introductionmentioning

confidence: 98%

See 1 more Smart Citation

Fast and efficient automatic memory management for GPUs using compiler-assisted runtime coherence scheme

Pai

Govindarajan

Thazhuthaveetil

2012

Proceedings of the 21st International Conference on Parallel Architectures and Compilation Techniques

View full text Add to dashboard Cite

Exploiting the performance potential of GPUs requires managing the data transfers to and from them efficiently which is an errorprone and tedious task. In this paper, we develop a software coherence mechanism to fully automate all data transfers between the CPU and GPU without any assistance from the programmer. Our mechanism uses compiler analysis to identify potential stale accesses and uses a runtime to initiate transfers as necessary. This allows us to avoid redundant transfers that are exhibited by all other existing automatic memory management proposals.We integrate our automatic memory manager into the X10 compiler and runtime, and find that it not only results in smaller and simpler programs, but also eliminates redundant memory transfers. Tested on eight programs ported from the Rodinia benchmark suite it achieves (i) a 1.06x speedup over hand-tuned manual memory management, and (ii) a 1.29x speedup over another recently proposed compiler-runtime automatic memory management system. Compared to other existing runtime-only and compiler-only proposals, it also transfers 2.2x to 13.3x less data on average.

show abstract