This week my guest is Shubho Sengupta, Research Scientist at Baidu. I had the pleasure of meeting Shubho at the Rework Deep Learning Summit earlier this year, where he delivered a presentation on Systems Challenges for Deep Learning. We dig into this topic in the interview, and discuss a variety of issues including network architecture, productionalization, operationalization and hardware.
Shubho has tons of insights to share into how to do deep learning at scale, and even if you're not operating at the scale of Baidu, I think you'll learn a lot from our conversation about architecting, productionalizing and operationalizing deep learning systems. We also spent some time discussing the role of GPUs and hardware in building scalable machine learning systems, an area that he has a lot to say about as the author of the CuDPP, the CUDA Data Parallel Primitives Library, which was the first parallel programming library for GPUs.