For more than four decades, transistors have been shrinking exponentially in size, and therefore the number of transistors in a single microelectronic chip has been increasing exponentially. Such an increase in packing density was made possible by continually shrinking the metal–oxide–semiconductor field-effect transistor (MOSFET). In the current generation of transistors, the transistor dimensions have shrunk to such an extent that the electrical characteristics of the device can be markedly degraded, making it unlikely that the exponential decrease in transistor size can continue. Recently, however, a new generation of MOSFETs, called multigate transistors, has emerged, and this multigate geometry will allow the continuing enhancement of computer performance into the next decade.

Multigate transistors as the future of classical metal–oxide–semiconductor field-effect transistors

This review paper explores considerations for ultimate CMOS transistor scaling Transistor architectures such as extremely thin silicon-on-insulator and FinFET (and related architectures such as TriGate, Omega-FET, Pi-Gate), as well as nanowire device architectures, are compared and contrasted Key technology challenges (such as advanced gate stacks, mobility, resistance, and capacitance) shared by all of the architectures will be discussed in relation to recent research results

Considerations for Ultimate CMOS Scaling

https://cyberleninka.org/article/n/1380904.pdf

ASAP7: A 7-nm finFET predictive process design kit

Predictive MOSFET models are critical for early stage design-technology co-optimization and circuit design research In this work, Predictive Technology Model files for sub-20nm multi-gate transistors have been developed (PTM-MG) Based on MOSFET scaling theory, the 2011 ITRS roadmap and early stage silicon data from published results, PTM for FinFET devices are generated for 5 technology nodes corresponding to the years 2012-2020 on the ITRS roadmap

Exploring sub-20nm FinFET design with predictive technology models

Compute-in-memory (CIM) is a new computing paradigm that addresses the memory-wall problem in hardware accelerator design for deep learning. The input vector and weight matrix multiplication, i.e., the multiply-and-accumulate (MAC) operation, could be performed in the analog domain within memory sub-array, leading to significant improvements in throughput and energy efficiency. Static random access memory (SRAM) and emerging non-volatile memories such as resistive random access memory (RRAM) are promising candidates to store the weights of deep neural network (DNN) models. In this review, firstly we survey the recent progresses in SRAM and RRAM based CIM macros that have been demonstrated in silicon. Then we discuss general design challenges of the CIM chips including analog-to-digital conversion (ADC) bottleneck, variations in analog compute, and device non-idealities. Next we introduce the DNN+NeuroSim benchmark framework that is capable of evaluating versatile device technologies for CIM inference and training performance from software/hardware co-design's perspective.

Compute-in-Memory Chips for Deep Learning: Recent Trends and Prospects

A high performance 22/20nm CMOS bulk FinFET achieves the best in-class N/P I on values of 1200/1100 µA/µm for I off =100nA/µm at 1V. Excellent device electrostatic control is demonstrated for gate length (L gate ) down to 20nm. Dual-Epitaxy and multiple stressors are essential to boost the device performance. Dual workfunction (WF) with an advanced High-K/Metal gate (HK/MG) stack is deployed in an integration-friendly CMOS process flow. This dual-WF approach provides excellent V th roll-off immunity in the short-channel regime that allows properly positioning the long-channel device V th . Enhanced 193nm immersion lithography has enabled the stringent requirements of the 22/20nm ground rules. Reliability of our advanced HK/MG stack is promising. Excellent SRAM static noise margin at 0.45V is reported.

High performance 22/20nm FinFET CMOS devices with advanced high-K/metal gate scheme

A leading edge 5nm CMOS platform technology has been defined and optimized for mobile and HPC applications. This industry-leading 5nm technology features, for the first time, full-fledged EUV, and high mobility channel (HMC) finFETs with densest 0.021µm2 HD SRAM. This true 5nm CMOS platform technology is a full node scaling from our successful 7nm node [4] in offering ~1.84x logic density, 15% speed gain or 30% power reduction. The 5nm platform technology successfully passed qualification [3] with consistently high yielding 256Mb HD/HC SRAM, and large logic test chip consisting of CPU/GPU/SoC blocks. Currently in risk production, this true 5nm platform technology is on schedule for high volume production in 1H 2020.

5nm CMOS Production Technology Platform featuring full-fledged EUV, and High Mobility Channel FinFETs with densest 0.021µm 2 SRAM cells for Mobile SoC and High Performance Computing Applications

C. P. Lin

Papers

High performance 22/20nm FinFET CMOS devices with advanced high-K/metal gate scheme

5nm CMOS Production Technology Platform featuring full-fledged EUV, and High Mobility Channel FinFETs with densest 0.021µm 2 SRAM cells for Mobile SoC and High Performance Computing Applications