Sessions

Computer Arithmetic Research to Accelerate Privacy-Protecting Encrypted Computing Such as Fully Homomorphic Encryption

Abstract

A major breakthrough of computer science in the 21st century has been the discovery and practical demonstration of encrypted computing technologies such as Fully Homomorphic Encryption (FHE). Encrypted computing technologies allow sensitive data to be encrypted such that arbitrary programs can be securely run over the encrypted data where the output, when decrypted, is equivalent to the result of running the original algorithm on the unencrypted data. In this talk we focus on the use of and potential for computer arithmetic research to enable practical encrypted computing, such as to accelerate advanced encryption implementations on custom hardware. These technologies are ground-breaking in their ability for privacy-preserving data science, machine learning and artificial intelligence on sensitive data sets with minimal costs in terms of engineering effort, power, compute resources, etc. We discuss theory, design, algorithmic, hardware, software engineering and systems research that is enabling applications of encrypted computing in regulated data industries, such as in medical and financial domains. We use our work using accelerating the PALISADE open-source homomorphic encryption software library for practical applications in case studies on the 3rd Generation Intel® Xeon® Scalable processor, using commercial GPUs, FPGA-based accelerators and custom ASIC designs. We discuss case studies from commercial applications.

Speaker

Kurt Rohloff, CTO and Co-founder, Duality Technologies

Dr. Kurt Rohloff is the CTO and Co-founder of Duality Technologies. He has been leading the development and application of practical Fully Homomorphic Encryption since it was first discovered in 2009. He is the co-founder of the PALISADE open-source library and co-founder of the HomomorphicEncryption.org industry consortium for FHE technologies. He has led multiple DARPA and IARPA efforts to develop and apply homomorphic encryption, and he has been awarded a DARPA Director's Fellowship and other recognition for his work on FHE. He has been a tenured professor at the New Jersey Institute of Technology and a senior scientist at Raytheon BBN Technologies. He received his undergrad degree from Georgia Tech and his MS and PhD from the Univ. of Michigan. https://www.linkedin.com/in/kurt-rohloff/