In this session I will start with identifying the current problems in Artificial Intelligence as data exists in most industries/edge in the form of isolated pockets with privacy constraints for sharing the data for training an AI model. While transferring data from edge storage to a centralized cloud for training -many problems arise. Performance issues, cost and security or regulatory compliance are some of them. I’ll then walk through detailed features of federated learning combined with Blockchain to solve the above mentioned problems . Next I'll explain the advantages of modern consortium blockchains which require decentralized data storage for validating transactions securely with examples of some of the decentralized data storage platforms across a global P2P network and some of the inherent problems related to storage size and distribution.I will explain the idea of combining AI and Blockchain for mutually benefiting solutions and how they overcome challenges in each - block chain to use for consensus and decision making in AI as well as AI for optimizing storage and overcoming challenges in Blockchain - with a careful grouping of storage/transactions . I'll also showcase some of the use cases with merits of the above approach to future digitization and automation - performance and security challenges.
AI & Blockchain - The Emerging Duets
For the past three decades, PCI-SIG® has delivered a succession of industry-leading PCI Express® (PCIe®) specifications that remain ahead of the increasing demand for a high-bandwidth, low-latency
We present RAINBLOCK, a public blockchain that achieves high transaction throughput without modifying the proof-of-work consensus.
AI/ML is not new, but innovations in ML models development have made it possible to process data at unprecedented speeds.
Modern AI systems usually require diverse data processing and feature engineering at a tremendous scale and employ heavy and complex deep learning model that requires expensive accelerators or GPUs
Compute, memory, storage, and connectivity demands are forcing the industry to adapt as it meets the expanding needs of cloud, edge, enterprise, 5G, and high-performance computing.
The next generation of automobiles moves to the adoption of PCIe for data communications in vehicles, and the JEDEC Automotive SSD solution enables a high performance, high reliability solution for
SNIA Developer Conference
September 16-18, 2024
SDC 2024 is brought to you by SNIA. SNIA is an industry association committed to its mission of worldwide leadership developing and promoting architectures, standards, education and vendor-neutral collaboration.
