Moore's Lobby’s Post

A data center the size of a shoebox? Imec’s Quentin Herr and Anna Herr explain in IEEE Spectrum how superconducting digital technology can manage the unsustainable energy requirements of computing, enable drastically smaller data centers and support AI and machine learning growth.   Learn about imec’s latest breakthroughs in engineering the basic superconducting materials, in circuit development and architectural design, and join us at #ITFWorld2024 (Antwerp, May 21 & 22), booth 19, to learn more about imec’s superconducting-based chip design. Read the full article: https://lnkd.in/eQSbpTsv

shoeboxDC-TM: According to a recent article in IEEE Spectrum, superconducting computer chips are showing promise due to their ability to easily remove the small amount of power that is dissipated. it is very challenging to stack computational chips on top of each other because of the large amount of power, and therefore heat, that is dissipated within the chips. In superconducting technology, the little power that is dissipated is easily removed by the liquid helium #SuperconductingTechnology #ComputationalChips #IEEEspectrum

HOW TO PUT A DATA CENTER IN A SHOEBOX One way to manage the unsustainable energy requirements of the computing sector is to fundamentally change the way we compute. Superconductors could let us do just that. Are the advantages enough to overcome the cost of cryogenic cooling? Our work suggests they most certainly are. As the scale of computing resources gets larger, the marginal cost of the cooling overhead gets smaller. Our research shows that starting at around 1016 floating-point operations per second (tens of petaflops) the superconducting computer handily becomes more power efficient than its classical cousin. This is exactly the scale of typical high-performance computers today, so the time for a superconducting supercomputer is now. https://lnkd.in/dtQt5f4i

AI is already driving up data center capacity: - 1. The global data center storage capacity is expected to grow from 10.1 zettabytes (ZB) in 2023 to 21.0 ZB in 2027. 2. The number of hyperscale data centers, which are essential for AI operations, doubled to 992 facilities by the end of 2023. This capacity is expected to double again in the next four years to meet the growing demand. 3. AI-related activities currently consume 10-20% of data center energy in the U.S., and this percentage is expected to increase significantly across all developed nations. The integration of quantum computing with AI is essential to revolutionize approaches to sustainability and scalability and make sure we can drastically reduce the energy consumption of computations, paving the way for more sustainable and cost-effective solutions. So its clearly a responsible thing to think beyond just the AI adoption hype and explore how to do this sustainably. This interview, gives some insight into what we are doing about that: https://okt.to/1d3hw9 #QuantumComputing #DrugDiscovery #Technology

Exploring Advances in Superconductor Computing with imec! imec USA is leading significant advancements in superconductor computing, which promises to enhance both computational speed and energy efficiency. Learn more about their innovative contributions by accessing the full article here: https://lnkd.in/eQSbpTsv

Check out our latest blog on the rise of neuromorphic systems and their impact on energy-efficient computing 🌟. Read more here: https://lnkd.in/eyxCvbJY

Breaking Moore’s Law: Lightmatter Accelerates Progress Toward Light-Speed Computing Lightmatter, founded by three MIT alumni, is using photonic technologies to reinvent how chips communicate and calculate. Our ability to cram ever-smaller transistors onto a chip has enabled today’s age of ubiquitous computing. But that approach is finally running into limits, with some experts declaring an end to Moore’s Law and a related principle, known as Dennard’s Scaling. Those developments couldn’t be coming at a worse time. Demand for computing power has skyrocketed in recent years thanks in large part to the rise of artificial intelligence, and it shows no signs of slowing down. Continue reading https://lnkd.in/dXyrbkSc

Emerging system requirements are driven by industry advancements in areas such as data centers, artificial intelligence, machine learning, and quantum computing, among others. Samtec 224 Gbps PAM4 interconnect systems are carefully designed to satisfy these high-performance requirements, as you’ll see in the video. https://lnkd.in/gJN9YKrt

🗞 Electronic News! 🗞 Rosini, the CEO of the company, is a seasoned entrepreneur with a background in founding companies in AI, gaming, and cellphone security. In a recent interview with eeNews Europe, Rosini expressed his motivation behind venturing into the development of a CPU architecture, considering it to be one of the most challenging tasks in the world. His drive to tackle this complexity stemmed from a desire to push boundaries and explore new frontiers in computing. On the other hand, Earley, who holds a PhD from the University of Cambridge, co-founded Vaire Computing in 2021 and assumed the role of CTO. Her research interests at Cambridge included reversible computing and molecular programming, which she is now translating into practical implementations at Vaire. The company's focus on reversible computing marks a departure from traditional computing paradigms that have dominated the industry for decades. #electricalengineering #electronics #embedded #embeddedsystems #electrical #computerchips Follow us on LinkedIn to get daily news: HardwareBee - Electronic News and Vendor Directory

Every week, I write the Future of Computing newsletter, featuring computing news, good reads, and my interviews with founders and industry experts 📧 Here are some key takeaways about what made the headlines in 2023: 1) Small language models started matching larger, closed-source models in performance while being small and (cost-)efficient enough to run locally. 2) We have strong signs that machine learning will help with writing poems and developing better materials, computer chips, or batteries – and assisting humans to shape the worlds of atoms. 3) Memory, not GPUs, has become the bottleneck holding back machine learning, so we need faster, high-bandwidth memory. 4) GPUs are better than CPUs for machine learning tasks, but people continued exploring what’s next – maybe chips leveraging in-memory computing or fully analog AI chips. 5) Similar in spirit, people are looking more than ever into what’s next beyond silicon – maybe another semiconductor material or even leveraging entirely new physical principles for computing – more on optical computing below. 6) Extreme ultraviolet lithography (EUV) came to fruition as TSMCs 3nm node reached volume production, among others, for Apple’s M3 chips series. 7) While larger and larger quantum computers were announced, we have seen in 2023 the first quantum processors with many logical qubits – an important milestone. 8) Access to quantum computers for experimentation has become fairly easy through cloud providers. Meanwhile, quantum-inspired, classical algorithms showed value in AI, e.g., for compressing machine learning models. 9) The photonics industry shifted its focus from optical AI accelerators to photonic interconnects as a potential remedy for the AI memory bottleneck. Subscribe to the newsletter and get access to a report diving deeper into what made the headlines in 2023👇 https://lnkd.in/esPNnA29