Deconstructing Deep Learning
As we enter the era of AI, new problems arise. Skilled data scientists are in short supply, data preparation is a time-suck, and a growing number of open source models conflate and confuse the ecosystem. Learn how a combination of next-gen hardware focused on widening I/O bottlenecks and accelerating across the stack, along with deep learning software focused on flattening the deep learning curve to value, can help overcome these difficulties.
Sumit Gupta, VP, HPC, AI & Machine Learning, IBM Systems