News on China's scientific and technological development.
- Thread starter Quickie
- Start date
Large-scale distributed diffractive-interference hybrid photonic chiplets
The rapid development of artificial general intelligence (AGI) introduces significant performance challenges for next-generation computing. Electronic devices such as graphics processing units (GPUs) are constrained by computational and energy efficiency limitations, hindering the advancement of modern AGI models. In contrast, photonic computing offers unprecedented low-power computing at the speed of light, promising superior performance for intelligent tasks., Spatial photonic computing, exemplified by diffractive deep neural networks (D2NNs),, achieves large-capacity computing but faces scalability issues due to the use of passive photonic devices. Meanwhile, integrated photonic computing, leveraging highly scalable Mach-Zehnder interferometers (MZIs), typically involves hundreds to thousands of parameters, posing challenges for large-capacity computing. Additionally, inherent analog noise and time-varying errors in these systems limit them to simple tasks and shallow models, which are inadequate for real-world AGI applications.
In their recently published paper, Xu et al. introduce Taichi, as illustrated in , a large-scale, highly scalable distributed photonic computing architecture designed for real-world AGI tasks, leveraging the advantages of optical diffraction and interference. The double diffractive units for large-scale input and output passively perceive high-dimensional data and compactly represent them through universal diffraction, as illustrated in . Task-specific feature embeddings are efficiently achieved via tunable matrix multiplication with fully reconfigurable MZI arrays. These components together form the scalable “DE-IE-DD” framework of Taichi, which can significantly reduce the required scale of the reconfigurable MZI array and support diverse and complex tasks with 3.8% reconfigurable part of 4256 total neurons. And the distributed architecture divides large tasks into several subtasks, which are parallel processed using Taichi chiplets, as depicted in . Computing resources are allocated to multiple independent clusters, each organized separately for subtasks and ultimately synthesized to handle complex advanced tasks. The authors report an experimental 1000-category classification of 91.89% on the 1623-category Omniglot dataset, marking the first attempt at achieving on-chip capability with 13.96 million neurons and an energy efficiency of 160 TOPS/W. This represents a highly promising approach in the field.
Shanghai Institute of Microsystem and Information Technology has realized an ultra-high-speed, photon-number-resolvable photon quantum detector.
Recently, the team of Li Hao and You Lixing from the Shanghai Institute of Microsystem and Information Technology, Chinese Academy of Sciences, used sandwich-structured superconducting nanowires and multi-wire parallel operation to realize an ultra-high-speed, photon-number-resolvable photon detector with a maximum counting rate of 5 GHz and a photon number resolution of 61. The related results were published online on May 31, 2024 in the first-tier academic journal Photonics Research of the Chinese Academy of Sciences under the title "Superconducting single photon detector with speed of 5 GHz and photon number resolution of 61" ().
Increased amount of work hours dont necessarily reflect same increase in productivity. Having time for yourself and family is critically important too.
I used to work at Panasonic as an engineer and the Japanese would work similarly to 996 but would spend 5 hours a day sleeping at the desk. Total waste of time as far as I could tell. I doubt 996 workers in China are exactly brimming with productivity for 12 hours straight either. Anyone who’s worked in engineering knows that productivity per hour worked follows a curve and that it goes straight down at a point.
and now they have alot of problems in society that can be traced to the work culture. Same for SK.
supercat
Major
New procedure to treat hypertrophic heart disease doesn't need open-heart surgery.
New brain imaging equipment:
New brain imaging equipment:
It can take high-resolution brain images.
broadsword
Brigadier
I hope China will develop a device to clear blocked arteries with one treatment without invasive surgery, thus obsoleting balloon angioplasty. Demand for such machines will be high and factory production lines will resemble dumpling factories to meet Western demand, especially in the US whose people can then continue to gorge on their 2500kcal lunch.