
Doron Zarchy*
28 posts



Fhenix has been working toward a single goal: VISA-level throughput for encrypted execution on Ethereum. For years, that sounded unrealistic. Today, our research team published DBFV - a new FHE scheme designed to make encrypted arithmetic meaningfully faster: eprint.iacr.org/2025/2321 Why this matters: addition and multiplication are the core building blocks behind AI inference and financial rails. Make those cheap, and entire new classes of encrypted applications stop being theoretical. This is an early step. But the direction is clear: we're making FHE scalable in a way previously unimagined Stay tuned - 2026 will be FHE coded.

Fhenix has been working toward a single goal: VISA-level throughput for encrypted execution on Ethereum. For years, that sounded unrealistic. Today, our research team published DBFV - a new FHE scheme designed to make encrypted arithmetic meaningfully faster: eprint.iacr.org/2025/2321 Why this matters: addition and multiplication are the core building blocks behind AI inference and financial rails. Make those cheap, and entire new classes of encrypted applications stop being theoretical. This is an early step. But the direction is clear: we're making FHE scalable in a way previously unimagined Stay tuned - 2026 will be FHE coded.












Tomorrow, 2:30 PM (Taiwan time) - Doron Zarchy, Fhenix Research, presents live at @acm_ccs, one of the most prestigious conferences in security and cryptography. Until now, threshold FHE decryption was the bottleneck. Now it’s 20 000× faster and 37× lower latency. Co-authored by @GuyZys, @ChrisPeikert, @DZeresh, and @max_zengo, this paper unlocks Private DeFi - and far beyond. x.com/i/broadcasts/1…







