The size of transistors has drastically reduced over the years. Interconnects have likewise also been scaled down. Today, conventional copper (Cu)-based interconnects face a significant impediment to further scaling since their electrical conductivity decreases at smaller dimensions, which also worsens the signal delay and energy consumption. As a result, alternative scalable materials such as semi-metals and 2D materials were being investigated as potential Cu replacements. In this paper, we experimentally showed that CoPt can provide better resistivity than Cu at thin dimensions and proposed hybrid poly-Si with a CoPt coating for local routing in standard cells for compactness. We evaluated the performance gain for DRAM/eDRAM, and area vs. performance trade-off for D-Flip-Flop (DFF) using hybrid poly-Si with a thin film of CoPt. We gained up to a 3-fold reduction in delay and a 15.6% reduction in cell area with the proposed hybrid interconnect. We also studied the system-level interconnect design using NbAs, a topological semi-metal with high electron mobility at the nanoscale, and demonstrated its advantages over Cu in terms of resistivity, propagation delay, and slew rate. Our simulations revealed that NbAs could reduce the propagation delay by up to 35.88%. We further evaluated the potential system-level performance gain for NbAs-based interconnects in cache memories and observed an instructions per cycle (IPC) improvement of up to 23.8%.
Quantum machine learning (QML) is a category of algorithms that employ variational quantum circuits (VQCs) to tackle machine learning tasks. Recent discoveries have shown that QML models can effectively generalize from limited training data samples. This capability has sparked increased interest in deploying these models to address practical, real-world challenges, resulting in the emergence of Quantum Machine Learning as a Service (QMLaaS). QMLaaS represents a hybrid model that utilizes both classical and quantum computing resources. Classical computers play a crucial role in this setup, handling initial pre-processing and subsequent post-processing of data to compensate for the current limitations of quantum hardware. Since this is a new area, very little work exists to paint the whole picture of QMLaaS in the context of known security threats in the domain of classical and quantum machine learning. This SoK paper is aimed to bridge this gap by outlining the complete QMLaaS workflow, which encompasses both the training and inference phases and highlighting significant security concerns involving untrusted classical or quantum providers. QML models contain several sensitive assets, such as the model architecture, training/testing data, encoding techniques, and trained parameters. Unauthorized access to these components could compromise the model's integrity and lead to intellectual property (IP) theft. We pinpoint the critical security issues that must be considered to pave the way for a secure QMLaaS deployment.
Cloud hosting of quantum machine learning (QML) models exposes them to a range of vulnerabilities, the most significant of which is the model stealing attack. In this study, we assess the efficacy of such attacks in the realm of quantum computing. We conducted comprehensive experiments on various datasets with multiple QML model architectures. Our findings revealed that model stealing attacks can produce clone models achieving up to $0.9\times$ and $0.99\times$ clone test accuracy when trained using Top-$1$ and Top-$k$ labels, respectively ($k:$ num\_classes). To defend against these attacks, we leverage the unique properties of current noisy hardware and perturb the victim model outputs and hinder the attacker's training process. In particular, we propose: 1) hardware variation-induced perturbation (HVIP) and 2) hardware and architecture variation-induced perturbation (HAVIP). Although noise and architectural variability can provide up to $\sim16\%$ output obfuscation, our comprehensive analysis revealed that models cloned under noisy conditions tend to be resilient, suffering little to no performance degradation due to such obfuscations. Despite limited success with our defense techniques, this outcome has led to an important discovery: QML models trained on noisy hardwares are naturally resistant to perturbation or obfuscation-based defenses or attacks.