The new method accelerates encrypted matrix multiplication, advancing practical fullyhomomorphic encryption (FHE) for AI
DESILO, a Privacy Enhancing Technology (PET) company, and Cornami, a leader in scalable compute acceleration, today announced new research that significantly improves the performance of encrypted AI computation using fully homomorphic encryption (FHE).
The research paper – co-authored by Craig Gentry, widely recognized as the father of FHE and a Gödel Prize laureate, and Yongwoo Lee, Head of Cryptography at DESILO – introduces a new method designed for efficient encrypted matrix arithmetic. According to results reported in the paper, the approach delivers up to 80× faster encrypted matrix multiplication compared to representative state-of-the-art baselines, marking a meaningful step toward in practical deployment of privacy-preserving AI.
Matrix multiplication is the computational backbone of modern machine learning models. The new method is designed to provide an optimized pathway for encrypted matrix multiplication across diverse scales and real-world workloads, narrowing the long-standing gap between theoretical cryptographic schemes and operational AI systems.
“This research shows that privacy-preserving computation can be both efficient and practical,”said Seungmyung Lee, CEO of DESILO. “It forms a critical foundation for our encrypted AI stack, which enables organizations to analyze sensitive data without exposing it.”
The work reflects the strategic collaboration between DESILO and Cornami, combining DESILO's advances in FHE-based computation with Cornami's high-performance compute architecture. The joint effort focuses on making FHE usable in enterprise and AI environments where both data confidentiality and computational efficiency are essential.
“For decades, Fully Homomorphic Encryption (FHE) has been the gold standard for data privacy, but its computational cost has made real-world use impractical,” saidDr. Craig Gentry, Chief Scientist of Algorithms atCornami.
Our latest research focuses onaccelerating encrypted matrix multiplication, a fundamental operation representing over 90% of AI workloads. Combined with Cornami'sscalable Compute Fabric, it deliversorders-of-magnitude fasterencrypted processing.”
“This makesprivacy-preserving AIpractical,” added Gentry. “Efficient, secure matrix multiplication enables Large Language Models to run inference on encrypted data with near-plaintext performance, strengthening compliance, sovereignty, and post-quantum security.”
The research paper is now publicly available via the IACR ePrint Archive.
About DESILODESILO develops encrypted computation and privacy-preserving AI infrastructure that enables multi-party data collaboration without exposing sensitive information. The company is expanding its presence in healthcare and finance, building multi-institutional collaboration models for areas such as credit scoring and multi-omics research.
About CornamiCornami is the developer of amassively parallel, next-generation compute fabricthat deliversscalable performancefor data-intensive workloads. Its architecture enables advanced encryption technologies – includingFully Homomorphic Encryption (FHE)- to run atpractical speeds, makingprivacy-preserving AI viable across industries that demand bothsecurity and performance scale.
Media Contact
DESILOmarketing@desilo.ai
Cornamimedia@cornami.com
https://edge.prnewswire.com/c/img/favicon.png?sn=CN21262&sd=2025-11-12
View original content:https://www.prnewswire.com/news-releases/desilo-and-cornami-announce-breakthrough-in-encrypted-ai-computation-302612271.html
SOURCE DESILO
https://rt.newswire.ca/rt.gif?NewsItemId=CN21262&Transmission_Id=202511121200PR_NEWS_USPR_____CN21262&DateId=20251112