Chip最新文献

Extreme ultraviolet lithography (EUVL) has been demonstrated to meet the industrial requirements of new-generation semiconductor fabrication. The development of high-power EUV sources is a long-term critical challenge to the implementation of EUVL in high-volume manufacturing (HVM), together with other technologies such as photoresist and mask. Historically, both theoretical studies and experiments have clearly indicated that the CO₂ laser-produced plasma (LPP) system is a promising solution for EUVL source, able to realize high conversion efficiency (CE) and output power. Currently, ASML's NXE:3400B EUV scanner configuring CO₂ LPP source system has been installed and operated at chipmaker customers. Meanwhile, other research teams have made different progresses in the development of LPP EUV sources. However, in their technologies, some critical areas need to be further improved to meet the requirements of 5 nm node and below. Critically needed improvements include higher laser power, stable droplet generation system and longer collector lifetime. In this paper, we describe the performance characteristics of the laser system, droplet generator and mirror collector for different EUV sources, and also the new development results.

Integrated photonic quantum circuits (IPQCs) have attracted increasing attention in recent years due to their widespread applications in quantum information science. While the most envisioned quantum technologies such as quantum communications, quantum computer and quantum simulations have placed a strict constraint on the scalability of chip-integrated quantum light sources. By introducing size-confined nanostructures or crystal imperfections, low-dimensional semiconductors have been broadly explored as chip-scale deterministic single-photon sources (SPSs). Thus far a variety of chip-integrated deterministic SPSs have been investigated across both monolithic and hybrid photonic platforms, including molecules, quantum dots, color centers and two-dimensional materials. With the rapid development of the chip-scale generation of single photons with deterministic quantum emitters, the field of IPQCs has raised new challenges and opportunities. In this paper, we highlight recent progress in the development of waveguide-coupled deterministic SPSs towards scalable IPQCs, and review the post-growth tuning techniques that are specifically developed to engineer the optical properties of these WG-coupled SPSs. Future prospects on stringent requirement for the quantum engineering toolbox in the burgeoning field of integrated photonics are also discussed.

The spike-response model (SRM) describes the adaptive behaviors of a biological neuron in response to repeated or prolonged stimulation, so that SRM neurons can avoid information overload and support neural networks for competitive learning. In this work, an artificial SRM neuron with the leaky integrate-and-fire (LIF) functions and the adaptive threshold is firstly implemented by the volatile memristive device of Pt/NbO_x/TiN. By modulating the volatile speed of the device, the threshold of the SRM neuron is adjusted to achieve the adaptive behaviors, such as the refractory period and the lateral inhibition. To demonstrate the function of the SRM neuron, a spiking neural network (SNN) is constructed with the SRM neurons and trained by the unsupervised learning rule, which successfully classifies letters with noises, while a similar SNN with LIF neurons fails. This work demonstrates that the SRM neuron not only emulates the adaptive behaviors of a biological neuron, but also enriches the functionality and unleashes the computational power of SNNs.

The emerging applications of composite gels as thermal interface materials (TIMs) for chip heat dissipation in intelligent vehicle and wearable devices require high thermal conductivity and remarkable damping properties. However, thermal conductivity and damping properties are usually correlated and coupled each other. Here, inspired by Maxwell theory and adhesion mechanism of gecko's setae, we present a strategy to fabricate polydimethylsiloxane-based composite gels integrating high thermal conductivity and remarkable damping properties over a broad frequency and temperature range. The multiple relaxation modes of dangling chains and the dynamic interaction between the dangling chains and aluminum fillers can efficiently dissipate the vibration energy, endowing the composite gels with ultrahigh damping property (tan δ > 0.3) over a broad frequency (0.01 – 100 Hz) and temperature range (–50 – 150 °C), which exceeds typical state-of-the-art damping materials. The dangling chains also comfort to the interfaces between polymer matrix and aluminum via van der Waals interaction, resulting in high thermal conductivity (4.72 ± 0.04 W m^–1 K^–1). Using the polydimethylsiloxane-based composite gel as TIMs, we demonstrate effective heat dissipation in chip operating under vigorous vibrations. We believe that our strategy could be applied to a wide range of composite gels and lead to the development of high-performance composite gels as TIMs for chip heat dissipation.

We present a theoretical investigation, based on the tight-binding Hamiltonian, of efficient second- and third-order nonlinear optical processes in the lattice-matched undoped ${\left(\text{GaP}\right)}_N/{\left(S{i}_2\right)}_M$ short-period superlattice that is waveguide-integrated in a microring resonator on an opto-electronic chip. The nonlinear superlattice structures are situated on the optically pumped input area of a heterogeneous “XOI” chip based on silicon. The spectra of ${\chi }_{zzz}^{\left(2\right)}\left(2\omega ,\omega ,\omega \right)$, ${\chi }_{xzx}^{\left(2\right)}\left(2\omega ,\omega ,\omega \right)$, ${\chi }_{xxxx}^{\left(3\right)}\left(3\omega ,\omega ,\omega ,\omega \right)$ and the Kerr refractive index ($n_2$), have been simulated as a function of the number of the atomic monolayers for “non-relaxed” heterointerfaces; These nonlinearities are induced by transitions between valence and conduction bands. The large obtained values make the ${\left(\text{GaP}\right)}_N/{\left(S{i}_2\right)}_M$ short-period superlattice a good candidate for future high-performance XOI photonic integrated chips that may include Si₃N₄ or SiC or AlGaAs or Si. Near or at the 810-nm and 1550-nm wavelengths, we have made detailed calculations of the efficiency of second- and third-harmonic generation as well as the performances of entangled photon-pair quantum sources that are based upon spontaneous parametric down conversion and spontaneous four-wave mixing. The results indicate that the ${\left(\text{GaP}\right)}_N/{\left(S{i}_2\right)}_M$ short-period superlattice is competitive with present technologies and is practical for classical and quantum applications.

Quantum computer, harnessing quantum superposition to boost a parallel computational power, promises to outperform its classical counterparts and offer an exponentially increased scaling. The term “quantum advantage” was proposed to mark the key point when people can solve a classically intractable problem by artificially controlling a quantum system in an unprecedented scale, even without error correction or known practical applications. Boson sampling, a problem about quantum evolutions of multi-photons on multimode photonic networks, as well as its variants, has been considered as a promising candidate to reach this milestone. However, the current photonic platforms suffer from the scaling problems, both in photon numbers and circuit modes. Here, we propose a new variant of the problem, membosonsampling, exploiting the scaling of the problem can be in principle extended to a large scale. We experimentally verify the scheme on a self-looped photonic chip inspired by memristor, and obtain multi-photon registrations up to 56-fold in 750,000 modes with a Hilbert space up to ${10}^{254}$. The results exhibit an integrated and cost-efficient shortcut stepping into the “quantum advantage” regime in a photonic system far beyond previous scenarios, and provide a scalable and controllable platform for quantum information processing.

In this paper, we review the recent trends in parallel search and artificial intelligence (AI) applications using emerging non-volatile ternary content addressable memory (TCAM). Firstly, the principle and development of four typical emerging memory used to implement the non-volatile TCAM are discussed. Then, we analyze the principle and challenges of SRAM-based TCAM and non-volatile TCAM for the parallel search. Finally, the research trends and challenges of non-volatile TCAM used for AI application are presented, which include computer-science oriented and neuroscience oriented computing.

In-memory computing has carried out calculations in situ within each memory unit and its main power consumption comes from data writing and erasing. Further improvements in the energy efficiency of in-memory computing require memory devices with sub-femto-Joule energy consumption. Floating gate memory devices based on two-dimensional (2D) material heterostructures have outstanding characteristics such as non-volatility, multi-bit storage, and low operation energy, suitable for application in in-memory computing chips. Here, we report a floating gate memory device based on a WSe₂/h-BN/Multilayer-graphene/h-BN heterostructure, the energy consumption of which is in sub-femto Joule (0.6 fJ) per operation for program/erase, and the read power consumption is in the tens of femto Watt (60 fW) range. We show a Hopfield neural network composed of WSe₂/h-BN/Multilayer-graphene/h-BN heterostructure floating gate memory devices, which can recall the original patterns from incorrect patterns. These results shed light on the development of future compact and energy-efficient hardware for in-memory computing systems.

Conventional von Neumann architecture faces many challenges in dealing with data-intensive artificial intelligence tasks efficiently due to huge amounts of data movement between physically separated data computing and storage units. Novel computing-in-memory (CIM) architecture implements data processing and storage in the same place, and thus can be much more energy-efficient than state-of-the-art von Neumann architecture. Compared with their counterparts, resistive random-access memory (RRAM)-based CIM systems could consume much less power and area when processing the same amount of data. In this paper, we first introduce the principles and challenges related to RRAM-based CIM systems. Then, recent works on the circuit and macro levels of RRAM-CIM systems will be reviewed to highlight the trends and challenges in this field.