In the past decade, there has been tremendous progress in integrating chalcogenide phase-change materials (PCMs) on the silicon photonic platform for non-volatile memory to neuromorphic in-memory computing applications. In particular, these non von Neumann computational elements and systems benefit from mass manufacturing of silicon photonic integrated circuits (PICs) on 8-inch wafers using a 130 nm complementary metal-oxide semiconductor line. Chip manufacturing based on deep-ultraviolet lithography and electron-beam lithography enables rapid prototyping of PICs, which can be integrated with high-quality PCMs based on the wafer-scale sputtering technique as a back-end-of-line process. In this article, we present an overview of recent advances in waveguide integrated PCM memory cells, functional devices, and neuromorphic systems, with an emphasis on fabrication and integration processes to attain state-of-the-art device performance. After a short overview of PCM based photonic devices, we discuss the materials properties of the functional layer as well as the progress on the light guiding layer, namely, the silicon and germanium waveguide platforms. Next, we discuss the cleanroom fabrication flow of waveguide devices integrated with thin films and nanowires, silicon waveguides and plasmonic microheaters for the electrothermal switching of PCMs and mixed-mode operation. Finally, the fabrication of photonic and photonic–electronic neuromorphic computing systems is reviewed. These systems consist of arrays of PCM memory elements for associative learning, matrix-vector multiplication, and pattern recognition. With large-scale integration, the neuromorphicphotonic computing paradigm holds the promise to outperform digital electronic accelerators by taking the advantages of ultra-high bandwidth, high speed, and energy-efficient operation in running machine learning algorithms.
Neuromorphic algorithms achieve remarkable performance milestones in tasks where humans have traditionally excelled. The breadth of data generated by these paradigms is, however, unsustainable by conventional computing chips. In-memory computing hardware aims to mimic biological neural networks and has emerged as a viable path in overcoming fundamental limitations of the von Neumann architecture. By eliminating the latency and energy losses associated with transferring data between the memory and central processing unit (CPU), these systems promise to improve on both speed and energy. Photonic implementations using on-chip, nonvolatile memories are particularly promising as they aim to deliver energy-efficient, high-speed, and high-density data processing within the photonic memory with the multiplexing advantages of optics. In this article, we overview recent progress in this direction that integrates phase-change material (PCM) memory elements with integrated optoelectronics. We compare performances of PCM devices using optoelectronic programming schemes and show that energy consumption can be significantly reduced to 60 pJ using picosecond (ps) optical pulse programming and plasmonic nanogap devices with a programming speed approaching 1 GHz. With these energy-efficient waveguide memories, concepts of in-memory photonic computing are implemented based on crossbar arrays. Compared with digital electronic accelerators: application-specific integrated circuits (ASICs) and graphics processing units (GPUs), photonic cores promise 1−3 orders higher compute density and energy efficiency, although much more work toward commercialization is still required.Graphical abstract
The ever‐increasing demands for data processing and storage will require seamless monolithic co‐integration of electronics and photonics. Phase‐change materials are uniquely suited to fulfill this function due to their dual electro‐optical sensitivity, nonvolatile retention properties, and fast switching dynamics. The extreme size disparity however between CMOS electronics and dielectric photonics inhibits the realization of efficient and compact electrically driven photonic switches, logic and routing elements. Here, the authors achieve an important milestone in harmonizing the two domains by demonstrating an electrically reconfigurable, ultra‐compact and nonvolatile memory that is optically accessible. The platform relies on localized heat, generated within a plasmonic structure; this uniquely allows for both optical and electrical readout signals to be interlocked with the material state of the PCM while still ensuring that the writing operation is electrically decoupled. Importantly, by miniaturization and effective thermal engineering, the authors achieve unprecedented energy efficiency, opening up a path towards low‐energy optoelectronic hardware for neuromorphic and in‐memory computing.
In this paper, we study the use of silicon microring modulators in an integrated mode-division multiplexing (MDM) network for short-reach interconnects using discrete multi-tone (DMT) modulation. A fiber grating coupler array is used as a high-speed optical interface between the on-chip mode channels and off-chip multicore fiber (MCF). The whole link has a crosstalk level lower than -23.1 dB. The MDM network can work without the crosstalk compensation of digital signal processing (DSP). We present experimental results of the linearity of the silicon microring modulator in terms of spurious free dynamic range (SFDR). The third order intermodulation distortion (IMD3) SFDR is 93.4 dB/Hz$^{2/3}$ at 20 GHz. We can improve system performance by using adaptive nonlinear equalizer at the DMT receiver to mitigate transmitter and transmission impairments. An aggregate bitrate above 312 Gb/s is achieved with a bit error ratio (BER) of 3.8 × 10$^{-3}$. With the addition of nonlinear equalizer, the line rates of the MDM network can be further increased to about 330 Gb/s. The experimental results demonstrate the feasibility of single wavelength DMT transmission in MDM network with MCF.
28 August 2019 in IEEE Journal of Selected Topics in Quantum Electronics
We present the first experimental demonstration at the optical telecommunication band of near-isotropic and transverse-electric (TE) photonic bandgap (PBG) in a hyperuniform disordered photonic structured (HUDPS) wall network. Next, we propose and demonstrate state-of-the-art performance of a hyperuniform disordered silicon photonic (HUDSiP) polarizer. The TE mode is blocked by the combined effects of reflection due to a wide PBG in the optimal wall network structure, scattering, and radiative bandedge resonances. The transverse-magnetic (TM) mode is allowed to pass through with a small insertion loss (IL). A 12.9-μm-long HUDSiP polarizer is experimentally demonstrated with an ultrabroad bandwidth of 240 nm (1440–1680 nm) when the extinction ratio (ER) ≥ 30 dB and IL ≤ 3 dB for the TM mode. The measured spectrally averaged ER and IL of the TM mode are 41.4 dB and 0.8 dB (0.4 dB at 1550 nm), respectively. A 38.7-μm-long HUDSiP polarizer shows an average IL for the TE mode of 54.9 dB and ER of 53.0 dB. Tests of 40-Gbps on-off-keying data transmission through the HUDSiP polarizers show clear eye diagrams. The HUDSiP polarizers are promising in polarization filtering in the large-scale polarization-sensitive/diverse silicon photonic circuits for high-speed optical interconnects.