Mining Text using GPU Clusters

• funded by: ORNL
• funding level: $17,003
• duration: 10/24/2008 - 07/30/2009

This work seeks to assess the viability of text mining on clusters composed of graphics processing units (GPUs). The proposed work covers development of optimized text search algorithms for GPUs and novel program paradigms for massive data processing requirements in GPUs. To this end, this work will leverage and build on previous research (at ORNL and NCSU) on:

• customizing text search for the Nvidia CUDA programming platform for GPUs and
• high-level prototyping of data-parallel programming models.

Publications:

• "GPU-Accelerated Text Mining" by Y. Zhang, F. Mueller, X. Cui and T. Potok in Workshop on Exploiting Parallelism using GPUs and other Hardware-Assisted Methods, Mar 2009.
• "A Programming Model for Massive Data Parallelism with Data Dependencies" by Y. Zhang, F. Mueller, X. Cui and T. Potok in Workshop on Programming Models for Emerging Architectures, Sep 2009.