TensorTEE: Unifying Heterogeneous TEE Granularity for Efficient Secure Collaborative Tensor Computing

TL;DR

TensorTEE introduces a unified tensor-granularity TEE for heterogeneous CPU-NPU systems, significantly improving secure collaborative tensor computing efficiency with minimal overhead, enabling practical secure LLM training.

Contribution

It proposes a novel unified tensor-granularity TEE that addresses memory and security challenges in heterogeneous CPU-NPU environments, enhancing performance and security for tensor workloads.

Findings

TensorTEE improves LLM training performance by 4.0x.

It incurs only 2.1% overhead compared to non-secure training.

The system enables direct data transfer without re-encryption.

Abstract

Heterogeneous collaborative computing with NPU and CPU has received widespread attention due to its substantial performance benefits. To ensure data confidentiality and integrity during computing, Trusted Execution Environments (TEE) is considered a promising solution because of its comparatively lower overhead. However, existing heterogeneous TEE designs are inefficient for collaborative computing due to fine and different memory granularities between CPU and NPU. 1) The cacheline granularity of CPU TEE intensifies memory pressure due to its extra memory access, and 2) the cacheline granularity MAC of NPU escalates the pressure on the limited memory storage. 3) Data transfer across heterogeneous enclaves relies on the transit of non-secure regions, resulting in cumbersome re-encryption and scheduling. To address these issues, we propose TensorTEE, a unified tensor-granularity heterogeneous TEE for efficient secure collaborative tensor computing. First, we virtually support tensor granularity in CPU TEE to eliminate the off-chip metadata access by detecting and maintaining tensor structures on-chip. Second, we propose tensor-granularity MAC management with predictive execution to avoid computational stalls while eliminating off-chip MAC storage and access. Moreover, based on the unified granularity, we enable direct data transfer without re-encryption and scheduling dilemmas. Our evaluation is built on enhanced Gem5 and a cycle-accurate NPU simulator. The results show that TensorTEE improves the performance of Large Language Model (LLM) training workloads by 4.0x compared to existing work and incurs only 2.1% overhead compared to non-secure training, offering a practical security assurance for LLM training.