Focuslight Technologies’ Post

Researchers' approach may protect quantum computers from attacks . Quantum computers, which can solve several complex problems exponentially faster than classical computers, are expected to improve artificial intelligence (AI) applications deployed in devices like autonomous vehicles; however, just like their predecessors, quantum computers are vulnerable to adversarial attacks. #TechTrends #TechInnovationsDaily #DigitalFrontiers #FutureTechInsights

Quantum computers, which can solve several complex problems exponentially faster than classical computers, are expected to improve artificial intelligence (AI) applications deployed in devices like autonomous vehicles; however, just like their predecessors, quantum computers are vulnerable to adversarial attacks.

Quantum computers, which can solve several complex problems exponentially faster than classical computers, are expected to improve artificial intelligence (AI) applications deployed in devices like autonomous vehicles; however, just like their predecessors, quantum computers are vulnerable to adversarial attacks.

This week's #innovatorspotlight is Professor Daniel Lidar, director of the University of Southern California Quantum Computing Center at the Information Science Institute. Lidar has been a Professor of Electrical Engineering and Chemistry since 2005. Before his time at USC, Lidar was a faculty member at the University of Toronto. Since 1995, he has worked in quantum computing, which focuses on harnessing the power of qubits to solve complex problems much faster than traditional computers. Read the article below to learn more about Professor Lidar's work with quantum computing research, and don't forget to follow The Circuit | News for more technology news.

For 25 years, on-board digital processing has been implemented using MCUs, DSPs, FPGAs and ASICs, offering increasing levels of performance, resources and throughput: all these technologies are based on conventional computing techniques and memory interfaces, however, traditional architectures are handicapping the demand for higher performance due to memory-bandwidth constraints and power consumption. Future applications require significantly more on-board processing on increasing amounts of data in real-time, autonomously and intelligently, for data-extensive applications, e.g. machine learning, space-domain awareness and AI-enabled, ultra-high-definition video processing. Associative Compute-In-Memory is based on parallel processing and high-speed SRAM technology, where data is stored and processed directly within the memory, eliminating the need to retrieve data from off-chip storage using complex interfaces. Associative Compute-In-Memory On-Board Processing stores information in a highly parallelized and distributed manner, allowing for simultaneous processing of multiple data elements leading to faster results. Information is searched based on its content or relationships, without the need for addressing, and this natural access makes it easier to explore and analyse complex data sets, discover patterns and derive insights without the constraints of traditional storage. By leveraging inherent parallelism with fast data access, complex computations can be performed in real-time to gain valuable insights quicker and make data-driven decisions faster!   Associative Compute-In-Memory will revolutionize on-board processing with a road-map to offer a staggering 420 TOPS of performance at 12W, enabling the next generation of intelligent, autonomous applications!   My latest blog article, Peta Ops, Associative Compute-In-Memory On-Board Processing: Coming to a Spacecraft in Your Orbit, has just been published on the Spacechips’ website: https://lnkd.in/eTFnhWEv   What you will learn:   1. Principles of Associative Compute-In-Memory On-Board Processing 2. Advantages of Associative Compute-In-Memory for SAR and LIDAR applications 3. How you can access the first devices for your next satellite application 4. Future, smart on-board processor for your next satellite application   Out-of-this-World-Design (https://lnkd.in/ey5RhUYK) is my monthly blog on Space Electronics and a free education resource used by the global space industry. For ten years, I have been posting articles on topics such as satellite-systems engineering, space-grade FPGAs, semiconductor radiation effects, spacecraft power generation, SEE mitigation, testing satellite payloads, PCB design for spacecraft avionics and satellite applications, as well as reviewing the latest components and technologies. Each month you can contribute to an on-line discussion with other readers on the chosen topic to improve your learning, and also participate in a quiz and win a prize!

#Topics Scientists Develop the First Universal, Programmable, and Multifunctional Photonic Chip [ad_1] A team from Universitat Politècnica de València and iPRONICS has developed the world’s first programmable and multifunctional photonic chip, enhancing data flow management in telecommunications and AI systems. (Artist’s concept). Credit: SciTechDaily.comA collaborative team from UPV and iPRONICS has developed the market’s first universal, programmable, and multifunctional photonic chip.Researchers from the Photonics Research Laboratory (PRL)-iTEAM at the Universitat Politècnica de València, in collaboration with iPRONICS, have developed a groundbreaking photonic chip. This chip is the world’s first to be universal, programmable, and multifunctional, making it a significant advancement for the telecommunications industry, data centers, and AI computing infrastructures. It is poised to enhance a variety of applications including 5G communications, quantum computing, data centers, artificial intelligence, satellites, drones, and autonomous vehicles.The development of this revolutionary chip is the main result of the European project UMWP-Chip, led by researcher José Capmany and funded by an ERC Advanced Grant from the European Research Council. The work has been publis...

Klepsydra AI and Frontgrade Gaisler Collaborate to Expand Microprocessing Versatility in Space Missions Through AI   Klepsydra AI, a leading provider of artificial intelligence (AI) software solutions, and Frontgrade Gaisler, a world leader in embedded computer systems for harsh environments, have announced an initiative to adapt AI software to run on microprocessors used in space missions. This effort is part of a contract recently awarded to Klepsydra AI by the European Space Agency (ESA) to port Klepsydra AI to Gaisler’s GR740 and NOEL-V processors, enabling high performance and reliable AI applications.   By successfully porting Klepsydra AI software to Frontgrade Gaisler’s space-grade GR740 and GR765 microprocessors, artificial intelligence algorithms can enhance the versatility of these microprocessors, enabling a wide range of AI-driven tasks in space, including remote sensing, Earth observation, data analysis, autonomous decision-making, and anomaly detection. The Klepsydra AI software will also be adapted for use with RTEMS, an open-source, real-time operating system widely used in space applications. https://lnkd.in/dbgKz7Fg #AIinSpace #SpaceExploration #InnovationInTech

#Topics Scientists Develop the First Universal, Programmable, and Multifunctional Photonic Chip [ad_1] A team from Universitat Politècnica de València and iPRONICS has developed the world’s first programmable and multifunctional photonic chip, enhancing data flow management in telecommunications and AI systems. (Artist’s concept). Credit: SciTechDaily.comA collaborative team from UPV and iPRONICS has developed the market’s first universal, programmable, and multifunctional photonic chip.Researchers from the Photonics Research Laboratory (PRL)-iTEAM at the Universitat Politècnica de València, in collaboration with iPRONICS, have developed a groundbreaking photonic chip. This chip is the world’s first to be universal, programmable, and multifunctional, making it a significant advancement for the telecommunications industry, data centers, and AI computing infrastructures. It is poised to enhance a variety of applications including 5G communications, quantum computing, data centers, artificial intelligence, satellites, drones, and autonomous vehicles.The development of this revolutionary chip is the main result of the European project UMWP-Chip, led by researcher José Capmany and funded by an ERC Advanced Grant from the European Research Council. The work has been publis...

After months of testing AI models, planning downstream tasks, and testing methodologies and tools on HPC4AI, we are ready to run on 4k GPUs for 5M GPU hours and 160TB of data; we are ready to start the training from scratch of our "neuromante" foundational model. Stay tuned for the scientific results. With Bruno Casella Gianluca Mittone Concetto Spampinato. Thanks to Leonardo supercomputing at CINECA. ICSC - Centro Nazionale di Ricerca in HPC, Big Data e Quantum Computing Università degli Studi di Torino Dipartimento di Informatica - Università degli Studi di Torino EuroHPC Joint Undertaking (EuroHPC JU) #hpc #hpc4ai #hpcai #highperformancecomputing #ai #aihealth

Can AI help us find extraterrestrial life?! NVIDIA joined hands with SETI Institute to use real-time data perception and machine learning using its GPUs to detect Fast Radio Bursts (FRBs) from outer space. The latest trial involved collecting over 90 billion packets of data from the Crab Nebula and processing them in just 15hrs for detection. The capability of AI systems to process data at scale can mean accelerated progress in the search for life outside of Earth. This is extremely cool, and certainly a push in the right direction for AI integration in human exploration for something bigger than us 🤩 🌌 #artificialintelligence #space #nvidia #gpu #ml #seti #dataanalytics #bigtech #bigdata #aiforgood