Cheng Qian
Ph.D., Principal Investigator
Email:qiancheng@@gdiist.cn
Personal Profile
Cheng Qian graduated from the Junior Class of University of Science and Technology of China in 2004 and the Institute of Computing Technology of Chinese Academy of Sciences in 2010 with a PhD degree. He has been engaged in the research and development of high-performance general purpose processor (CPU) and neural Network processor (NPU) projects in Loongson Institute of Computing Technology, Cambrian Team and Shanghai Center for Brain Science and Brain-Inspired Technology. Dr.Qian has rich engineering experience in the field of VLSI design, and has been granted more than 11 invention patents and published 8 SCI papers. Since 2016, he has mainly worked on the industrialization of deep learning processors, serving as the vice president of Cambrian Technology Corporation. He has been mainly responsible for the ecology of intelligent computing systems, the ecology of intelligent processor chips, and university cooperation. He has established the Cambrian R&D department and government affairs system, and organized teams to apply for and undertake more than 50 research projects. In recent years, he has presided over and participated in 4 important scientific research projects in the field of integrated circuits and more than 10 chip design-related projects. Since 2019, he has been the director and chief engineer of Brain Inspired Computing System of Brain and Intelligence Science and Technology Research Institute of Zhang Jiang Laboratory. He has received the Special Award for Scientific and Technological Achievement Transformation from the Beijing Branch of Chinese Academy of Sciences, and the 2019 Chinese Academy of Sciences Outstanding Scientific and Technological Achievement Award for "Deep Learning Processor Architecture Research" (collective).He has held social positions as deputy director of the Special Committee of the National University Artificial Intelligence Big Data Alliance, Standing member of the Expert Committee of the China Big Data and Intelligent Computing Industry Alliance, member of the Financial Committee of the Western Returned Scholars Association, and expert database of the Ministry of Science and Technology.
Laboratory of Cognitive Computing and Decision-Making System
The group is mainly engaged in the research of cognitive computing and decision-making system(CCDMS). CCDMS refers to a new type of computing that mimics the functioning of the human brain with the goal of more accurate models of how the human brain/mind senses, reasons, and responds to stimulus. Inspired by the brain neural structure and information processing mechanism, the research group deeply cross-integrates artificial intelligence with neuroscience, cognitive science, psychology and other disciplines, and relies on the intelligent computing power of Hengqin Advanced Intelligent Computing Center to develop a set of distributed computing framework, algorithm engine, tools and simulation platform. Our current research focuses on generative AI and embodied intelligence, such as the diffusion model and transformer-based reinforcement learning. Our group’s long-term vision is to breaking through the existing intelligence models in cognitive decision-making, autonomous learning, interactive learning, adaptive, swarm intelligence and other aspects, and realizing the next generation of brain-like computing systems with 'mechanism like brain, behavior like human', so that machines can engage in more intellectual and creative work, and help humans make efficient decisions.
Representative publications
Papers:
（1）Chen L, Cong M, Huang J, Cheng Q, et al. A novel hardware/software partitioning for SIMD-based real-time AVS video decode, Multimedia Tools and Applications, 2014, 71(3):1651-1671.
（2）Qian Cheng, Shen Haihua, Chen Tianshi, and Chen Yunji,Survey of Design-for-Debug of VLSI, 2012, 49(1):21-34.
（3）Qian Chen, Liu Daofu, Chen Yunji,A NOC trace compression method for multi-core processors 2011, 21(3):254-260.
（4）Hu W, Chen Y, Chen T, Cheng Q，et al. Linear Time Memory Consistency Verification，IEEE Transactions on Computers, 2011, 61(4):502-516.
（5）Li L, Chen T, Chen Y, Cheng Q, et al. Brief announcement: program regularization in verifying memory consistency， Proceedings of the, ACM Symposium on Parallelism in Algorithms and Architectures, SPAA 2011，San Jose, Ca, Usa, 2011.6.4-6.7
Patents:
（1）CN201610979814- Operation apparatus and method for acceleration chip for accelerating deep neural network algorithm；
（2）CN201510824901- Pipeline data synchronization apparatus and method for multi-input multi-output processor；
（3）CN201510863726- Adder device, data accumulation method, and data processing device
（4）CN201510862723- Data accumulation apparatus and method, and digital signal processing device；
（5）CN201510825061- Pipeline-level operation device, data processing method and network-on-chip chip；
（6）CN200910237056- Device and system for realizing debuggability of multicore processor EJTAG；
（7）CN201210082890- Entropy decoding device and method for realizing CAVLC (context-based adaptive variable length coding) of H.246
（8）CN200810247389- Clock-gating system and operating method thereof
（9）CN200910238032- Control device and control method of DMA controller
（10）CN200910237057- Method containing four instructions and supporting fast Fourier transformation operation
