The video discusses a new photonic chip that combines high-frequency radio waves and light for computing. It claims to be 1,000 times faster and 400 times more power-efficient than conventional processors. The chip is based on lithium niobate, a transparent material that can integrate photonic circuits.
The technology mixes a terahertz laser (light carrier) with a microwave signal to encode information, performing operations like differentiation and image recognition using optical components. It has an ultra-broad processing bandwidth of 67 GHz, much higher than 5G communication.
However, the paper overlooks the power consumption of the laser, waveform generator, and readout components. The chip can only perform limited operations like differentiation and relies on conventional computers for tasks like matrix multiplication and neural network inference.
The author praises the attempt to combine different domains but feels the technology is more useful for data communication than computing at this stage. The author sees photonics and eventually quantum computing as potential milestones for achieving higher computing efficiency.
Integrated microwave photonics (MWP) is an intriguing field that leverages integrated photonic technologies for the generation, transmission, and manipulation of microwave signals in chip-scale optical systems 1,2. In particular, ultrafast processing and computation of analog electronic signals in the optical domain with high fidelity and low latency could enable a variety of applications such as MWP filters 3–5 , microwave signal processing 6–9 , and image recognition 10,11 . An ideal photonic platform for achieving these integrated MWP processing tasks shall simultaneously offer an efficient, linear and high-speed electro-optic (EO) modulation block to faithfully perform microwave-optic conversion at low power, and a low-loss functional photonic network that can be configured for a variety of signal processing tasks, as well as large-scale, lowcost manufacturability to monolithically integrate the two building blocks on the same chip. In this work, we demonstrate such an integrated MWP processing engine based on a thin-film lithium niobate (LN) platform capable of performing multi-purpose processing and computation tasks of analog signals up to 256 giga samples per second (GSa/s) at CMOS-compatible voltages. By integrating a high-speed EO modulation block and a multi-purpose low-loss signal processing section on the same chip fabricated from a 4-inch wafer-scale process, we demonstrate high-speed analog computation, i.e., first- and second-order temporal integration and differentiation with processing bandwidths up to 67 GHz and computation accuracies up to 98.0 %, and deploy these functions to showcase three proof-of-concept applications, namely, ordinary differential equation (ODE) solving, ultra-wideband (UWB) signal generation and high-speed edge detection of images. We further leverage the image edge detector to enable a photonic-assisted image segmentation model that could effectively outline the boundaries of melanoma lesion in medical diagnostic images, achieving orders of magnitude faster processing speed and lower energy consumption than conventional electronic processors. Our ultrafast LN MWP engine could provide compact, lowpower, low-latency, and cost-effective solutions for future wireless communications, Internet of things, high-resolution radar systems and photonic artificial intelligence.
