US DARPA Promotes JUMP (Joint University Microelectronics Program) to Focus on Artificial Intelligence (AI), Autonomic Systems, Cellular Technologies, Nanotechnology, Networking, Hardware and Software Design and Memory, Breaking Existing Electronic System Performance and Scalability Limits , To promote the development of microelectronics-related technologies and innovative solutions to meet the future needs of the US Department of Defense (DoD), national security agencies and commercial companies in 2025-2030. According to The Next Platform Report, November 2016 DARPA incorporates AMD , Analog Devices, Arm, IBM, Intel, Micron, Samsung, TSMC and other semiconductor companies, and EMD Performance Materials, Lockheed-Martin, Northrop Grumman, Raytheon and other companies, for a five-year budget of 200 million US dollars, focusing on the field of microelectronics JUMP, DARPA and other companies each contribute 40% and 60% respectively DARPA and non-profit semiconductor research firm SRC (Semiconductor Research Corp.) Screening of proposals from various agencies in 2017, announced in January 2018 Establishment of 6 research centers , Covering more than 30 researchers at various universities in the US SRC responsible for management pointed out that DARPA hopes that the combination of public and private sectors JUMP military and business The application provides a wide range of applicability, including the creation of technical advantages for DoD in advanced radar, communications and weapons and other fields in the United States. It greatly enhances the technology of soldiers sensing the environment, processing data and communication capabilities, and drives the development of commercial IT in the industrial community. ADA Center (Applications Driving Architectures Center) hopes to completely transform the current design method, significantly reduce the complexity, cost, effort and technical barriers to the establishment and development of future advanced computing systems, expand the talent pool and promote creative ideas. Researchers will build models. Grouped design and manufacturing process, creating a system of software and hardware plug-and-play ecosystem. ASCENT (Applications and Systems-driven Center for Energy-Efficient Integrated Nanotechnologies) aims to surpass the limits of CMOS technology such as Moore's Law, researchers will focus on Innovative device technologies, integration solutions and hardware accelerator applications to improve performance, efficiency, energy consumption and data transfer bottlenecks for future computing systems such as high performance computing (HPC) and enterprise systems, continuous use of AMD and NVIDIA GPUs, or Intel and Xilinx field programmable logic gate arrays (FPGA), Good system performance and energy consumption. The goal of C-BRIC (Center for Brain-inspired Computing Enabling Autonomous Intelligence) is to surpass existing machine learning applications and lay the foundation for a new generation of autonomous intelligent systems for autonomous flying drones and personal intelligent assistants. Personnel will explore neuroscience algorithms and theories, hardware architecture, application drivers, and develop new semiconductor technologies to meet future system performance and energy consumption requirements. ComSecTer (Center for Converged Tera Hertz Communications and Sensing) will develop future mobile network infrastructure. Technology needed to support the data needs of technologies such as ultra-high-resolution imaging, graded precision positioning, lightweight whisper radio, etc. Researchers can use these techniques to solve some of the current car-related communication, navigation and security challenges and Since the revolution in the car and the rise of smart roads, IoT has included edge devices, networks and the cloud, and the upload of cloud-processed big data has put pressure on existing networks. CONIX (Computing On Network Infrastructure for Pervasive Perception, Cognition, and Action) is targeted at the edge device The development of a networked computing infrastructure between the cloud, the ability to move processing and decisions out of the cloud through intelligence built into the network, and create greater flexibility for existing and future IoT applications. CRISP (Center for Research on Intelligence Storage Research and processing-in-memory researchers will focus on eliminating the barriers between storage and storage, affecting the speed and efficiency of big data access, and the fact that computer systems have reached the bottleneck of memory wall technology for 70 years, hindering big data in technology. Applications for development. Researchers will build computing power into chip-level memory and integrate processor and memory chips into a 3D stack, which will facilitate the use of big data drives innovations in areas such as medical research and national security. DARPA Has always been committed to promoting the development of emerging technologies that are beneficial to the military and business. In 2016, DARPA sponsored the Cyber Grand Challenge held at the DefCon Security Conference to find the best autonomous network security solution. In 2017, Georgia Tech was selected, Intel, Northrop Grumman, Pacific Northwest National Lab, Qualcomm for HIVE (Hierarchical Identify Verify Exp Loit) plans to develop graphic analysis processors that can improve the efficiency of big data analysis and strengthen US companies' competitiveness and national security. 2017 DARPA and Brown University, Columbia University, Fondation Voir et Entrendre, John B. Pierce Lab, Paradromics, University of California Berkeley and others signed a partnership to conduct a neuro-engineering system (NESD) program to create a communication interface between the human brain and the digital world by converting electro-chemical signals from the brain's neurons into 0/1 digital information that drives the computer program. Development of implantable computing systems. DARPA also provided multi-million dollar funding to national laboratories, universities, NVIDIA and Rex Computing, etc. to develop new CPU and GPU technologies for the construction of next-generation computing systems. Is a thousand times the super computer.
