Computational Modeling

BES: Energy-Efficient Intelligent Computing with Emerging Microelectronics

Daniel Morton — Thu, 09 Apr 2026 19:47:10 +0000

BES: Energy-Efficient Intelligent Computing with Emerging Microelectronics Daniel Morton Thu, 04/09/2026 - 13:47

Event Details

Thursday April 9, 2026 | 3:00 - 4:00 PM

SEEC Sievers Room (S228)

Abstract:

The pursuit of high-performance, energy-efficient artificial intelligence (AI) opens exciting opportunities for emerging semiconductor memories and unconventional architectures. To maximize the potential of emerging computing technologies, innovations across the stack (from devices to architecture) become critical. In this talk, I will present our recent hardware-software co-design efforts in exploiting beyond-complementary-metal-oxide-silicon (CMOS) microelectronics for developing efficient deep neural network (DNN) hardware accelerators. We will investigate how to co-design the emerging microelectronics and crossbar architecture to address two major challenges of compute-in-memory: the analog functional errors and the peripheral overhead of analog-to-digital conversion. Furthermore, we will exploit the device physics of emerging non-volatile memory to enable stochastic and approximate computation, thereby achieving an optimal trade-off between efficiency and accuracy for AI inference. Our device-to-system co-optimization demonstrates exciting opportunities for beyond-CMOS microelectronics in developing the next-generation efficient compute systems for edge sensing and precision agriculture applications.

Biography:

Cheng Wang is an Assistant Professor of Electrical and Computer Engineering at Iowa State University. Cheng received his B.S. degree in Physics from Peking University in 2009 and completed his Ph.D. at the University of Texas at Austin in 2016, with his dissertation on emerging non-volatile and spintronic memories. Prior to joining Iowa State, Cheng was a Research Scientist at the Center for Brain-inspired Computing Enabling (C-BRIC) at Purdue University. Cheng worked as a Staff R&D Engineer at Seagate Research Center from 2016 to 2019, where he designed high-density magneto-electronic memory and storage technologies. His current research interests include machine learning hardware acceleration and energy-efficient neuromorphic computing with emerging technologies and architectures. He has served on the Technical Program Committee for beyond-CMOS and emerging technologies for IEEE/ACM Design Automation Conference (DAC), International Conference on Computer-Aided Design (ICCAD), and Great Lakes Symposium on VLSI. He is a recipient of the NSF CAREER Award, Seagate FRC Technical Award, and Best Paper Awards for the IEEE International Conference on Rebooting Computing (ICRC) and IEEE Cross-disciplinary Conference on Memory-Centric Computing.

April 2026

Off

Traditional

White

BES: Atoms to Bits: Toward Thermodynamic Intelligence

Daniel Morton — Thu, 02 Apr 2026 19:26:24 +0000

BES: Atoms to Bits: Toward Thermodynamic Intelligence Daniel Morton Thu, 04/02/2026 - 13:26

Event Details

Thursday April 2, 2026

SEEC Building N224 | 3:00 - 4:00 PM

Abstract:

As transistors approach atomic limits, heat dissipation has become the defining constraint of modern computing. My research asks a different question: can heat itself be used to compute? Using correlated quantum materials such as vanadium dioxide (VO2), we explore how electronic phase transitions driven by the flow of heat and charge, generate nonlinear dynamics that resemble neural behavior.

I will discuss our recent experiments revealing spiking, synchronization, memory, and stochasticity in Mott oscillators, as well as collective switching in thermally coupled device networks. These studies uncover how local phase transition fluctuations and mesoscale heat transport give rise to emergent order and functional computation. By linking atomic-scale phase transitions to network-level information processing, this work outlines a physical pathway from atoms to bits, pointing toward a thermodynamic framework for intelligent, energy-aware electronics.

Biography:

Daniel Morton — Tue, 09 Dec 2025 16:00:00 +0000

The Grid’s New Shock Absorber: ‘Droop-e’ Control tames frequency swings and keeps renewable energy flowing smoothly Daniel Morton Tue, 12/09/2025 - 09:00

Daniel Morton

Find out more

Check out the article

Electricity is crucial to modern life. We rely on being able to plug devices in to the outlet in the wall, flipping a switch, and things working without a problem. But it is not that simple, the power grid, all of the infrastructure that delivers energy from the power plant to your home, is something of a balancing act.

In order for a grid to operate safely, supply must equal demand. The flow of electricity through the grid in the United States flows at a frequency of 60 Hz, if the supply increases more than the demand, the frequency will increase, while if the demand increases, or the supply dips, the frequency decreases. The vast array of hardware that makes up the grid and in electrical devices, such as transformers, motors, or electronics, have been designed to operate at a specific frequency. If the grid is unbalanced, and the frequency changes too much, equipment can be damaged, efficiency is reduced, and it can lead to overheating, system failures, and blackouts. Keeping the grid online, and safe, is a balancing act that requires sophisticated controls systems to make sure that supply always equals demand.

Think of the electric grid like a high-speed train system. In order for the train system to operate effectively all the trains need to maintain consistent speeds and keep to schedule, so passengers are not left waiting on the platform, or miss their trains because they left too early. Traditional power plants are like massive freight trains, that are super heavy and take a long time to speed up or slow down. These massive freight trains provide a kind of inertia to the whole system. They are hard to disrupt, which results in a consistent speed. Renewable energy sources, such as solar and wind, are more like fast, light commuter trains, that can change speed essentially instantly. They lack the inertia of the massive freight trains, but they can change fast. If it were up to just human conductors and train line controllers to regulate how the trains are running, having the freight trains and the commuter trains on the same lines would be near impossible, the difference in speed and inertia would make it really hard to reconcile. This is where advanced computer-driven control systems come into play. In the train analogy the smart predictive system would predictively control the brake and throttle of the commuter trains to ensure a simple constant speed. How would the grid be impacted if we developed a smart control system?

This new report details work led by RASEI Fellow Bri-Mathias Hodge, and discloses a new approach for smart control in the grid. The grid is already full of control systems, with the standard way power generators respond to frequency events being via linear droop control. This would be like a simplistic cruise control for one of the Commuter Trains in the above analogy. If the frequency drops a little, the system increases the power proportionately. The problem with this approach is that it often doesn’t use the inverters full capacity fast enough. The innovation described in this work is an update called Droop-e, a non-linear control based on an exponential function. Think of it like replacing the trains cruise control, which previously had a simple on/off switch, with a smart responsive gas pedal, that can speed up, or slow down, on a curve.

This change, from an on/off control to a responsive curve, has the potential to have significant impacts on the grid. By using the available power reserves more effectively, Droop-e reduces the number of severe power swings in the system, and results in a slower rate of change of frequency (ROCOF), which can buy grid operators valuable time to react to changes in frequency.

The benefits from the ‘shock-absorber’ properties that Droop-e offers could help prevent blackouts before they start, help stabilize the grid and improve integration of renewable energy sources, and create a smarter, more responsive grid, future-proofing systems by replacing the software, and not the hardware, a significant cost saving.

The simulations from this study confirm that this new control approach could improve the stability of grids that include a combination of traditional power plants and renewable energy generators. If a major power plant trips offline, this sophisticated control system activates un-tapped power reserves in batteries and renewables, acting as a hyper responsive shock absorber to protect the entire grid system. For grid operators, it means more time to react. For everyone with devices plugged into a power outlet, it means improved stability, and less equipment damage. It also provides a more effective mechanism to integrate different power sources, improving reliability, security, and affordability.

December 2025

Off

Traditional

White