Is anybody aware of researches around Big Data and use of GPUs(exploiting their parallelism) to solve related(Big Data) challenges? I'm just curious, could you enlist some criteria which may be used(again, about Big Data issues); any suggestions would be greatly appreciated;
Dear Hassan, please take a look at these papers:
1) http://dl.acm.org/citation.cfm?id=1531668
2) http://link.springer.com/chapter/10.1007/978-3-642-36318-4_6#page-1
3) http://hgpu.org/?p=11734
- A variable, Yee-like grid, on which electric and magnetic fields are approximated.
- Particle-Pushers following the works of Boris and Vay
- Current-Deposition schemes published by Esirkepov and Villasenor-Buneman
- Macro-Particle Form Factors: NGP, CIC, TSC, PSQ
- Parallel HDF5 Output
https://www.hzdr.de/db/Cms?pNid=3227
Data DATA SCIENCE - WHEREWITHAL Application in Agriculture and Ma...
Hell Hassan,Big Data such as his name sais, it is the analysis of BIG quantities of Data, it doesn't care the specific area.
GPUs, FPGAs, Intel Xeon Phi, Parallella and other parallel architectures are exploiting their parallelism to solve problem in many areas. Bioinformatic, Text mining (Social Network, Internet), image (cryptography, etc...) , sound (speech recognition), video (real time face and object recognition), health cares, etc... So, for its efforts in energy consumption GPGPU (green500.org) has been continuously increasing in several scientific fields.
MapReduce and Hadoop are specific platforms to analysis different datasets exploting the parallelism of different machine connected for a network (Clusters and parallel computing). These platforms are developing and implementing tool for GPUs http://wiki.apache.org/hadoop/CUDA%20On%20Hadoop and http://www.cs.kent.edu/~zwang/schedule/dc8.pdf, for example.
NVIDIA has launched some APIs for Machine Learning (http://www.nvidia.com/object/machine-learning.html) and little months ago https://developer.nvidia.com/cuDNN.
Nowadays, we can find large dataset open to the public, look this link http://www.quora.com/Where-can-I-find-large-datasets-open-to-the-public.
Finally, I humbly would suggest that you select different dataset and explore them with MapReduce or Hadoop (of my own preference). There are some process that you have to do, HDF5 is a ver good format files to analyse dataset in tool for Big Data, and very much knowledge to acquire.
I hope this will help you, cheers,
---
Marcos
Dear Hassan, please take a look at these papers:
1) http://dl.acm.org/citation.cfm?id=1531668
2) http://link.springer.com/chapter/10.1007/978-3-642-36318-4_6#page-1
3) http://hgpu.org/?p=11734
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Algorithms