20 May 2021 in IEEE Transactions on Circuits and Systems I: Regular Papers
This paper presents a charge-domain in-memory computing (IMC) macro for precision-scalable deep neural network accelerators. The proposed Dual-SRAM cell structure with coupling capacitors enables charge-domain multiply and accumulate (MAC) operation with variable-precision signed weights. Unlike prior charge-domain IMC macros that only support binary neural networks or digitally compute weighted sums for MAC operation with multi-bit weights, the proposed macro implements analog weighted sums for energy-efficient bit-scalable MAC operations with a novel series-coupled merging scheme. A test chip with a 16-kb SRAM macro is fabricated in 28-nm FDSOI process, and the measured macro throughput is 125.2-876.5 GOPS for weight bit-precision varying from 2 to 8. The macro also achieves energy efficiency ranging from 18.4 TOPS/W for 8-b weight to 119.2 TOPS/W for 2-b weight.
13 April 2022 in IEEE Transactions on Circuits and Systems I: Regular Papers
A novel Resistive random access memory (ReRAM)-based Compute-in-memory (CIM) macro is proposed to overcome the limited accuracy and throughput of a conventional ReRAM-based CIM macro that results from to the low R-Ratio and large variation of ReRAM. The proposed structure consists of 1T2R1C bit-cells and 4-kb ReRAM-based nvCIM architecture with ternary weight and ternary input. Ternary multiplication is implemented with voltage division between paired ReRAM devices within a bit-cell to make the output voltage variation tolerant and less sensitive to low R-ratios. An accumulation operation is realized with capacitive coupling so that linearity can be guaranteed for a large number of operands, allowing accurate and fast multiply-and-accumulate (MAC) operations. For comprehensive validation of the proposed CIM macro, the Verilog-A models for ReRAM devices with an adjustable R-ratio and adjustable variations are adopted to perform simulation on various R-ratio and variation conditions. With the peripheral circuits designed in 180-nm CMOS technology, the proposed CIM macro is confirmed to have high variation tolerance, high throughput, and less sensitivity to a low R-ratio, resulting in a high ternary DNN accuracy of 99.07% (0.01% drop) for the MNIST and 83.79% (0.38% drop) for the CIFAR-10 data sets with an R-ratio as low as 38 and 20%/40% low/high resistance variation.
29 January 2024 in IEEE Transactions on Circuits and Systems II: Express Briefs
This article introduces an area-efficient SRAM-based Physically Unclonable Function (PUF) designed using a conventional 6T-SRAM architecture, thereby also facilitating its use as SRAM. By exploiting read current mismatch induced by the variation in the access transistor’s threshold voltage, a key is generated by comparing the voltages of bitlines discharged from two adjacent SRAMs. Using the maximum mismatch detection stabilization technique, the PUF’s stability is enhanced by identifying the access transistors exhibiting the highest read current mismatch in two adjacent SRAM cells. The proposed PUF is fabricated in 28-nm FDSOI process and has a bitcell area of 328F2. Although under nominal conditions, the native bit error rate (BER) and unstable bits ratio (UBR) are 1.92% and 16.44%, respectively, the proposed stabilization technique and the application of 7-bit temporal majority voting improved the BER and UBR to 0.056% and 0.37%, respectively.
1 January 2026 in IEEE Transactions on Information Forensics and Security