Abstract:Low power system-on-chips (SoCs) are now at the heart of Internet-of-Things (IoT) devices, which are well known for their bursty workloads and limited energy storage - usually in the form of tiny batteries. To ensure battery lifetime, DVFS has become an essential technique in such SoC chips. With continuously decreasing supply level, noise margins in these devices are already being squeezed. During DVFS transition, large current that accompanies the clock speed transition runs into or out of clock networks in a few clock cycles, and induces large Ldi/dt noise, thereby stressing the power delivery network (PDN). Due to the limited area and cost target, adding additional decap to mitigate such noise is usually challenging. A common approach is to gradually introduce/remove the additional clock cycles to increase or reduce the clock frequency in steps, a.k.a., clock skipping. However, such a technique may increase DVFS transition time, and still cannot guarantee minimal noise. In this work, we propose a new noise-aware DVFS sequence optimization technique by formulating a mixed 0/1 programming to resolve the problems of clock skipping sequence optimization. Moreover, the method is also extended to schedule extensive wake-up activities on different clock domains for the same purpose. The results show that we are able to achieve minimal-noise sequence within desired transition time with 53% noise reduction and save more than 15-17% power compared with the traditional approach.

An Adaptive Wide-Voltage-Range Droop Detection and Protection System Assisted with Timing Error Detection in 28nm CMOS

Noise-aware DVFS Transition Sequence Optimization for Battery-Powered IoT Devices.

A Timing Failure Tolerance Design with In-Field Simultaneous Error Detection and Correction.

Eliminating Timing Errors Through Collaborative Design to Maximize the Throughput

Negative Design Margin Realization through Deep Path Activity Detection Combined with Dynamic Voltage Scaling in a 55 nm Near-Threshold 32-Bit Microcontroller

A Fully Synthesizable Capacitorless Digital LDO for Distributed Power Delivery Network

Margin Elimination Through Timing Error Detection in a Near-Threshold Enabled 32-bit Microcontroller in 40-nm CMOS

25.8 A Near- Threshold-Voltage Network-on-Chip with a Metastability Error Detection and Correction Technique for Supporting a Quad-Voltage/Frequency-Domain Ultra-Low-Power System-on-a-Chip.

An Impedance Adapting Compensation Scheme for High Current NMOS LDO Design

A Capacitor-less Digital LDO Using Ripple-Frequency-Adaptive Time-domain Digital Pre-distortion Technique

Adaptive Design Techniques for High-speed Toggle 2.0 NAND Flash Interface Considering Dynamic Internal Voltage Fluctuations

14.7 A Modular Hybrid LDO with Fast Load-Transient Response and Programmable PSRR in 14nm CMOS Featuring Dynamic Clamp Tuning and Time-Constant Compensation

Edsu: Error Detection And Sampling Unified Flip-Flop With Ultra-Low Overhead

A Robust Asynchronous Approach for Realizing Ultra-Low Power Digital Self-Adaptive VDD Scaling System

An Energy-Efficient Microprocessor Using Multilevel Error Correction For Timing Error Tolerance

An error-resilient wavelet-based ECG processor under voltage overscaling

A Near Threshold Error Resilient Processor Based on Dynamic Timing Error Prediction and Within-a-cycle Timing Error Correction

Timing error resilient clock gate design for wide-voltage application

A Wideband High PSRR Capacitor-Less LDO with Adaptive DC Level Shift and Bulk-Driven Feed-Forward Techniques in 28nm CMOS

Margin Elimination in a 55 nm Near-Threshold Microcontroller with Adaptive Prediction Capability and Voltage Scaling