Turning data into useful information is fundamentally about sussing out the relationships between phenomena, items, or things. One of the ways it’s done is via Graph Analytics—representing datasets as graphs, setting algorithms loose on them, and getting impressively directional results.
Examples, you ask? Frankly they’re innumerable, but here’s a short list offered by Tim Mattson, Senior Principal Engineer in Intel’s Parallel Computing Laboratory and the subject of this interview:
- Recommendation engines
- Web searches
- Power grid optimization and troubleshooting
- Road networks/maps (How do driving apps know where the traffic jams are? Ta-da!)
- Global police/security surveillance of phones, cameras, the Internet
- DNA/gene regulation
All driven and informed by the power of Graph Analytics.
Intrigued?
Then watch this Tech.Decoded chat with Tim and get a front row seat to this technology, including how it’s evolving, Intel’s approach working with the software/open source ecosystem to optimize it, and where it’s going in this new decade.
Do it. [13:17 mins]
Get the software
- Download Intel® Math Kernel Library and Intel® Data Analytics Acceleration Library—part of five FREE Intel® Performance Libraries that help you deliver impressive performance on Intel® architecture.
- SuiteSparse:GraphBLAS—a full implementation of the GraphBLAS standard.
More resources
Read these related blogs: