
The next wave of AI depends as much on systems, memory, and standards as it does on models. In this episode, Pankaj Mehra, Founder of Elephance Memory and a professor at The Ohio State University, connects the dots between early neural networks, modern AI infrastructure, and the standards decisions that will determine what actually gets adopted at scale.
Pankaj makes the case that developers will not tolerate dozens of incompatible heterogeneous computing APIs. That’s why he’s pushing for a single, usable interface across accelerators and smart devices, informed by work in Open Compute Project communities and the broader PyTorch ecosystem.
We close with a discussion on computational storage and computational memory, data gravity, and the sustainability math that’s forcing data centers to rethink where compute happens and how much energy gets burned moving bits instead of processing them.
SNIA is an industry organization that develops global standards and delivers vendor-neutral education on technologies related to data. In these interviews, SNIA experts on data cover a wide range of topics on both established and emerging technologies.
About SNIA:
Więcej odcinków z kanału "SNIA Experts on Data"



Nie przegap odcinka z kanału “SNIA Experts on Data”! Subskrybuj bezpłatnie w aplikacji GetPodcast.








