Haitong Li

The Spring issue of IEEE Solid-State Circuits Magazine is out! https://lnkd.in/gxbjNtKw This issue is dedicated to Integrated Circuits for AI and ML applications, and was guest edited by Marian Verhelst and Charlotte Frenkel. It features four excellent articles that highlight core challenges and methodologies in AI hardware acceleration (see https://lnkd.in/gayteYmV for more detail). The feature articles are: 1) "Design and Optimization of Efficient Digital Machine Learning Accelerators" by Wei Tang, Sung-Gun Cho, Jie-Fang Zhang, and Zhengya Zhang examines AI acceleration through optimized computational kernels, such as matrix–matrix multiplication, leveraging architectures like SIMD and systolic arrays (10.1109/MSSC.2025.3549361). 2) "SRAM- and eDRAM-Based Compute-in-Memory Designs, Accelerators, and Evaluation Frameworks," by Yifan He, Xiaofeng Hu, Hongyang Jia, and Jae-sun Seo addresses in-memory computing (IMC) as a means to reduce data movement overhead, covering both analog and digital IMC approaches (https://lnkd.in/gHiukWKx). 3) "Open Source Heterogeneous SoCs for Artificial Intelligence," by Francesco Conti, Angelo Garofalo, Davide Rossi, Giuseppe Tagliavini, and Luca Benini explores the integration of heterogeneous systems on chip, demonstrating how open source platforms such as PULP support scalable AI hardware development (https://lnkd.in/gAQgK-8d). 4) " Sparsity-Aware Hardware," by Man Shi, Adrian Kneip, Nicolas Chauvaux, Jiacong Sun, Charlotte Frenkel, and Marian Verhelst discusses sparsity-aware hardware design, illustrating how structured and unstructured sparsity can be leveraged for significant performance improvements using an adapted roofline model (https://lnkd.in/gv8eTyvd). The issue also features regular columns: 1. Behzad Razavi’s “The Analog Mind” column (https://lnkd.in/gxKyqZr7) continues the series on CMOS inverter applications 2. Chris Mangelsdorf’s “Shop Talk: What You Didn’t Learn in School” column picks up where the previous issue left off, continuing the discussion on Miller compensation in "Miller’s Wrong Half-Plane Zero" (https://lnkd.in/gngTrywV). This time, the focus is on the right half-plane zero and its potential to degrade stability if not properly addressed. 3. Farhana S.’s “Circuits From a Systems Perspective” traces the historical evolution of wireless charging in everyday devices (https://lnkd.in/gnVhqUyF). Lot's of interesting articles to read! Enjoy! #ieeesscc #analog #ieee #ai #ml #solidstatecircuits #hardwareacceleration #semiconductors #inmemorycomputing

260

3 Comments

Check out our latest work published in IEEE Electron Device Letters. In this paper, we introduce NeuroSPICE, a transient circuit simulator built on a Physics-Informed Neural Network (PINN) framework. We examine the feasibility of using PINNs to solve circuit equations and assess their strengths and limitations. While PINNs are unlikely to replace conventional SPICE solvers, their differentiable and flexible framework opens promising opportunities for inverse design, circuit optimization, and emerging device modeling beyond traditional SPICE inference. https://lnkd.in/gHa2YRUh #PhysicsInformedNeuralNetworks #CircuitSimulation #DeviceModeling

77

For someone working at the intersection of VLSI, AI hardware, and drones, the emphasis on convergence domains (AI with health, environment, and sustainability) validates the push we have been making toward application‑centric, system‑level thinking rather than siloed subjects. The stress on ethical reasoning and creative expression also resonates with our mentoring style, where we push students to think about impact, not just implementation. This shift gives us a strong policy cover to propose exactly the kind of modules we already run informally: research‑driven projects, FPGA/SoC design studios, and cross‑listed courses with agriculture, health tech, or robotics. It is also an opportunity to formalize “special topics” that track fast‑moving areas (e.g., neuromorphic sensing, autonomous UAV swarms, AI accelerators) and to demand the same flexibility in non‑IIT institutes by citing IIT practice and NEP 2020.

112

A thoughtful discussion initiated by a pioneer in wireless communication research, Prof. Robert Heath .The post and insightful comments by several world experts on the topic “How to grade homework in the GenAI era” will be extremely useful for many faculty members teaching engineering courses.

6

#MATLAB / #GNUOctave users, please check out #jdict - a tiny, universally-supported class that allows users to define attributes for any data arrays in both MATLAB (R2010+) and Octave (6.4+). https://lnkd.in/eRDpgnwh #Python users can use libraries like #xarray (https://lnkd.in/ecVVPeBm) to attach arbitrary attributes to an ND array, such as setting the labels of each dimension, and using such intuitive labels to index/slice the array. This has not been possible on MATLAB. With #jdict, one can define customized attributes to an array inside a hierarchical struct-like object, and allow you to easily navigate, set, add, retrieve these attributes just like what #xarray enables. It also allow one to directly use the dimension labels to access/select specific dimension of an array, similar to xarray.isel(). To use jdict is extremely simple, you only need to download a 500-line jdict.m unit from here and add it to your project https://lnkd.in/eyqSHSHK Please also check out this attached meme I made to compare jdict with struct, containers.Map and dictionary. I also attached an updated cheat-sheet to see all key syntaxes jdict supports. give it a try!

40

2 Comments