YAVIN: A Unified Architecture for Secure Edge Processing in Memory

2026-08-13

AI Summary

Existing trusted execution environments (TEEs) protect only in-processor computation and struggle to meet the multi-tenant security requirements of edge computing involving processors, memory, and accelerators. This work proposes a unified trusted computing base architecture that extends TEEs to dedicated protected memory regions, enabling secure decryption, processing, and re-encryption of encrypted data on untrusted memory buses by either the processor or in-memory compute units. For the first time in in-memory computing, it integrates the LightSaber post-quantum key encapsulation mechanism and ASCON-128 authenticated encryption, enhanced with DRAM-aware co-design and bit-sliced data ordering to enable secure and efficient tensor operations. Experiments demonstrate over 20× speedup compared to the state-of-the-art PIM-AES scheme and incur only 34% and 9.3% performance overhead on INT8 and INT32 quantized edge large models, respectively.

Abstract

Secure, private multi-tenant execution spanning processors, memory, and accelerators remains one of the most significant challenges in modern edge computing systems. Simultaneously, processing-in-memory (PIM) has emerged as an effective approach for reducing the Von Neumann bottleneck by moving computation closer to data. Existing trusted execution environments (TEEs) establish trust only within the processor, protecting data while it traverses untrusted resources such as the memory bus. Consequently, trusted computation cannot be performed directly within memory. We present YAVIN, a unified trusted computing base (TCB) that extends the TEE beyond the processor to encompass both processor execution and a dedicated memory region supporting trusted processing-in-memory execution while treating the memory bus as untrusted. Leveraging the dedicated protected memory regions already established by conventional TEE architectures, YAVIN enables data to be decrypted, processed, and re-encrypted by either processor or PIM execution while remaining within the TEE. To realize this unified TCB, YAVIN presents the first PIM implementations of the LightSaber KEM post-quantum cryptosystem and ASCON-128 authenticated encryption, co-designing both algorithms for efficient DRAM execution to establish and maintain shared cryptographic state. Finally, we demonstrate how cryptography-PIM co-design for tensor-based workloads reorganizes computation to satisfy the ordering constraints imposed by authenticated encryption with minimal performance overhead while simultaneously enabling bit-sliced ordering that limits temporary plaintext exposure. Compared to the latest PIM AES implementation, YAVIN achieves more than a 20x speedup while incurring only 34% and 9.3% overhead when executing INT8 and INT32 quantized edge-class LLMs, respectively, relative to plaintext execution.

Problem

• trusted execution environment
• processing-in-memory
• edge computing
• secure multi-tenant execution
• memory security

Innovation

• Processing-in-Memory (PIM)
• Trusted Execution Environment (TEE)
• Post-quantum Cryptography
• Co-design
• Edge Computing