Computation-in-memory (CIM) accelerators offer energy and area efficiency gains but remain vulnerable to power side-channel leakage. This paper presents a comprehensive security analysis of a pipelined digital SRAM-based CIM accelerator named DREAM-CIM. We first identify leakage mechanisms through pre-silicon modeling and gate-level simulation, and then confirm their presence on a 22 nm post-silicon prototype. Moreover, to mitigate leakage, we introduce a Globally Synchronous and Locally Asynchronous (GSLA) countermeasure approach that desynchronizes immediate-pipeline-update registers and randomizes local activity while preserving functional correctness. Security is evaluated using post-silicon power traces under standard side-channel analyses, demonstrating effective leakage suppression. The countermeasure incurs ≈ 53% area and ≈ 74% energy overhead relative to the baseline design. Compared to masking-based countermeasures, GSLA achieves a reduction of ≈ 8.9 × and ≈ 9.2 × in area and energy overhead, respectively. ... Read more

Compute-in-Memory (CiM) devices have several attributes that make them of interest especially for accelerating artificial intelligence workloads and, at the same time, reduce power consumption. Both are critical aspects of edge computing. This paper summarizes a 3-topic special session held at the European Test Symposium 2026 on the topic of CiMs: (a) concerning the reliability of CiMs, (b) the testing of CiMs, and finally (c) on the security of CiMs. ... Read more

Efficient generation of high-quality Gaussian random numbers remains a key challenge for enabling differential privacy (DP) and security primitives on resource-constrained edge devices. Existing hardware-based solutions rely on lookup tables or complex transformations, making them ill-suited for tightly integrated computation-in-memory (CIM) architectures. This work proposes a Gaussian random number generator (GRNG) based on a digital 8T SRAM-CIM array that leverages intrinsic memory periphery to generate and transform entropy. The design exploits voltage underscaling and stochastic readpath failures to produce in-situ uniform randomness, which is then aggregated through a lightweight central limit theorem (CLT)-based datapath with digital recentering to synthesize N0, σ2) samples. By reusing existing periphery, the proposed approach eliminates the need for dedicated random number generation hardware. Experimental evaluation demonstrates that the generated distribution closely follows an ideal normal curve, achieving low tail error and strong goodness-of-fit across standard statistical tests. The design sustains ≈ 2 Gsamples s throughput at an energy of ≈ 11.2 fJ/ sample, with no additional area overhead beyond standard CIM periphery. ... Read more

Mapping Binary Neural Networks (BNNs) on computation-in-memory (CIM) architectures enables a highly efficient approach for energy-constrained edge computing. In-memory processing significantly reduces critical performance bottlenecks in conventional architectures. Despite their efficiency, current optimized CIM implementations remain vulnerable to IP theft via side-channel analysis. This work investigates the side-channel leakage of a digital BNN-CIM accelerator that employs popcount-based accumulation. A range of circuit-level modifications in counter implementations are proposed and evaluated, exploring their impact on security metrics and design overhead. Results demonstrate that the Hamming weight (HW) and Hamming distance (HD) equalizing techniques combined with power equalization through duplication perform better than traditional dual-rail countermeasures. The findings provide practical guidance for designing secure and efficient peripheral components for popcount-based BNN accelerators. ... Read more

Binary Neural Networks (BNNs) have obtained a strong foothold in the field of machine learning at the edge due to their minimal hardware requirements. However, their energy and performance efficiency remain hindered by frequent data transfer between memory and processors. Computation-in-memory (CIM) architectures address this problem by embedding processing units within the memory. Unfortunately, current implementations of CIM are susceptible to IP piracy attacks through side channels. This paper presents a novel secure periphery scheme for NN accelerators with sequential accumulation that conceals IP information by obscuring the power consumption of the counter responsible for the leakage. This is achieved by combining two innovative techniques: operand schedule randomization and an always-count Gray code counter. The results demonstrate that the proposed design effectively resists power side channel attacks (SCAs). Moreover, Signal-to-Noise Ratio (SNR) and Test Vector Leakage Assessment (TVLA) show safe leakage levels. Compared to the state-of-the-art, our countermeasure reduces area and power overheads by up to 12.7× and 13.3×, achieving only 37% area and 51.2% power overhead with the added protection logic. Notably, this enhanced security comes with zero latency overhead, maintaining the performance of the baseline design. ... Read more

Computation-in-Memory (CIM) architectures address the rising demand for energy-efficient artificial intelligence (AI) solutions, by minimizing costly data movements between memory and processor. Within such architectures, SRAM-based digital CIM is especially attractive as it preserves the advantages of CIM while avoiding analog complexity. Recent studies have revealed potential weaknesses in these architectures, particularly to power side-channel attacks (SCA) capable of extracting sensitive model parameters (e.g., neural network (NN) weights), which represent the intellectual property of CIM-based neural network systems. In this study, we propose and evaluate two countermeasures to secure SRAM-based CIM architectures against power attacks: (1) Balanced Obfuscated-path countermeasure, and (2) Glitch Aware countermeasure. To validate their effectiveness, we conducted a comprehensive power analysis that successfully demonstrated attacks against an unprotected implementation. Our experimental results demonstrate that both countermeasures significantly improve resistance to power attacks. Although the Balanced Obfuscated-path offers better area overhead and run-time performance, the Glitch Aware approach achieves higher protection against advanced attacks, making each suitable for different design constraints. ... Read more

Artificial Intelligence (AI) has had a profound impact on our contemporary society, and it is indisputable that it will continue to play a significant role in the future. To further enhance AI experience and performance, a transition from large-scale server applications towards AI-powered edge devices is inevitable. In fact, current projections indicate that the market for Smart Edge Processors (SEPs) will grow beyond 70 Billion USD by 2026 [1]. Such a shift comes with major challenges, as these devices have limited computing and energy resources yet need to be highly performant. Additionally, security mechanisms need to be implemented to protect against diverse attack vectors as attackers now have physical access to the device. Besides cryptographic keys, Intellectual Property (IP), including neural network weights, may also be potential targets. The CONVOLVE [2] project (currently in its intermediate stage) follows a holistic approach to address these challenges and establish the EU in a leading position in embedded, ultra-low-power and secure processors for edge computing. It encompasses novel hardware technologies, end-to-end integrated workflows, and a security-by-design approach. This paper highlights the security aspects of future edge-AI processors by illustrating challenges encountered in CONVOLVE, the solutions we pursue including some early results, and directions for future research. ... Read more

Binary Neural Networks (BNNs) have demonstrated significant advantages in reducing computation and memory costs, all while maintaining acceptable accuracy on various image detection tasks. Thus, BNNs have the potential to support practical cognitive tasks on resource-constrained platforms, such as edge computing devices. To realize this, SRAM-based digital Computation-in-Memory (CIM) has gained growing attention as it overcomes the analog CIM architecture bottlenecks such as limited computing accuracy due to process variation, non-linearity, power and area-hungry Analog-to-Digital Converters (ADCs), etc. However, digital CIM architectures are highly dominated by power-hungry adder-trees, which can nullify the benefits of SRAM-based digital CIM. To address this issue, this paper proposes an adder free SRAM-based digital CIM, AFSRAM-CIM, for BNN acceleration. The proposed CIM architecture utilizes a multi-functional 10-T SRAM cell-based crossbar array and a new energy-efficient approach to perform the popcount operation. Simulation results using the MNIST dataset show that the proposed architecture maintains the state-of-the-art inference accuracy of 99.21% with only 11.86 fJ energy per operation. Moreover, AFSRAM-CIM achieves over 3× energy and ≈17× area savings when compared to the conventional digital CIM approaches. ... Read more

Computation-in-Memory (CIM) architectures present a promising solution for efficient implementation of Neural Networks. Particularly, SRAM-based digital CIM architectures are optimal candidates to realize them. Recent studies have revealed potential weaknesses in these architectures, particularly against power attacks. This study introduces a novel attack method enabling weight extraction through the analysis of the adder tree component within the architecture. In our attack, the k-means clustering technique is employed to identify the hamming weights of the CIM weights. Subsequently, we correlate traces belonging to known weights with traces belonging to Hamming groups with unknown weights in order to identify their weight values. As a case study, the attack was applied on SRAM CIM implementation based on 40nm TSMC technology. The results indicate that the weights stored in the CIM crossbar can be retrieved with 100% accuracy purely by analyzing the power consumption. ... Read more