Moving data is the biggest problem in computing. Many companies try to do all computation from the memory of one set of devices due to the cost of bandwidth, latency, and energy, as well as the act of running machine learning algorithms against one database instead of multiple ones. And if the company is still operating off the lagging speed of CPU-driven architectures, moving data could take anywhere from hours to days.
Enter the GPU Open Analytics Initiative (GOAI) and its first project, the GPU Data Frame (GDF). Enabled by NVIDIA’s hardware innovation providing 100x more processing cores and 20x greater memory bandwidth, the GDF is an Apache Arrow-based API that enables end-to-end GPU analytics by allowing for seamless passing of data between processes running on the same GPUs. This efficient data interchange improves performance, encouraging development of more sophisticated and interactive GPU-based applications.
Todd Mostak debuts the GDF, showcases the advanced speed and efficiency of GPUs, and highlights the importance of the open source community to enable efficient intra-GPU communication between different processes running on the GPUs. Todd explains in detail how the integration allows developers, data scientists, and researchers to build new functions to cluster or perform analysis on queries and seamless workflows that combine data processing, machine learning, and visualization.
This session is sponsored by MapD.
Todd Mostak is the founder of MapD. He is a graduate of Harvard’s Kennedy School of Government.
©2017, O'Reilly Media, Inc. • (800) 889-8969 or (707) 827-7019 • Monday-Friday 7:30am-5pm PT • All trademarks and registered trademarks appearing on oreilly.com are the property of their respective owners. • firstname.lastname@example.org