OptiPrime: Optimizing Private Inference through Protocol-Hardware Co-design
OptiPrime co-designs HE-MPC protocols with hardware acceleration to remove network communication bottlenecks in private DNN inference, beating Cheetah by up to 5.7x.
OptiPrime is a protocol-hardware co-optimization framework for private deep neural network inference based on hybrid homomorphic encryption (HE) and multi-party computation (MPC). It introduces a novel HE protocol for convolutions that reduces the number of transmitted output ciphertexts, addressing the network bottleneck that limits gains from commercial HE accelerators. A lightweight compression system reduces weight plaintext memory traffic by 10x, while a specialized dataflow maximizes on-chip reuse of intermediate ciphertexts. Experiments show up to 5.7x speedup over the Cheetah baseline on CPUs and 4.2x with an accelerator.
- Novel HE convolution protocol cuts transmitted output ciphertexts in HE-MPC private inference
- Commercial HE accelerators alone gave limited end-to-end gains due to network communication bottleneck
- Weight plaintext compression reduces memory traffic by 10 times
- Outperforms Cheetah baseline by up to 5.7x on CPUs and 4.2x with an accelerator
From the original article
1 min read at arxiv.orgPrivate deep neural network (DNN) inference based on hybrid homomorphic encryption (HE) and multi-party computation (MPC) can protect user data with a formal guarantee, but at the cost of significant latency overhead due to HE. Customized HE accelerators have been proposed and have achieved orders-of-magnitude speedup for individual HE operations.
By Jiangrui Yu · arxiv.org
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.