Francesca Rossi

#Best_Universities to #Pursue #Generative_AI in 2025, Choosing the right #university is crucial. This article explores some of the best universities to pursue generative AI #studies in 2025, #focusing on their #programs, #faculty, #research_opportunities, and #unique_offerings. https://lnkd.in/gYgtYrKY

3

1 Comment

This new artificial intelligence unit (#AIU), developed in collaboration with #IBM_Research, will enable future IBM Z systems to perform #AI_inferencing at a #massive_scale.

6

#Generative_AI and foundation models have already begun to revolutionize how the world lives and works. A recent report from #Stanford_University's Center for #Research on #Foundation_Models showed that #IBM's_model scored a #perfect #100% in several categories including #data_sources, #training compute resources, #harmful_data_filtration, model components, #terms_of_service, and #licensing details.

7

How does an organization harness the power of Generative AI at a large scale and not create a negative impact on the Environment ? (https://lnkd.in/dDCwp37v) Here comes the new AI chip from IBM. A compact 2U research prototype system using the IBM AIU NorthPole inference chip capable of processing 30k tokens/sec and below 1ms/token latency.   NorthPole is optimized for energy-efficiency and low latency (two conflicting goals).  NorthPole (in 12nm) can provide 70 times better energy efficiency (tokens/second/W) compared to 4nm GPU and ~50 times better latency (ms/token) compared to a 5nm GPU. #SustainableAI #Sustainability #GenerativeAI https://lnkd.in/gW8Pg9fM

20

Free virtual event coming on October 1st. Learn about industry trends and innovation spanning #AI, Hybrid #Cloud, and #Quantum-Safe #Security,

7

I really enjoyed this conversation as it allowed us to explore the fascinating world of Small Language Models and uncover why they're becoming the true workhorses of enterprise GenAI tasks! 💡 Given how frequently these questions come up in my discussions, I thought I'd share some key insights: 1️⃣ Why are SLMs important for enterprises? 🏢 Enterprises are leveraging LLMs and other GenAI solutions to improve their business processes, aiming to boost efficiency without inflating costs. While larger LLMs offer impressive capabilities, they come with significant cost implications - especially the proprietary models. Enter Small Language Models: though they might not match their giant cousins in raw performance initially, they can catch-up (and often outperform) through customization. Here's the game-changer: a task-customized SLM can deliver similar or superior performance compared to proprietary LLMs while being orders of magnitude more cost-effective! 💪 2️⃣ What is LLM customization? 🎯 To customize your LLM, you'll need two essential tools: - A solution to transform your enterprise data (knowledge) into training data and/or generate task-specific (skills) training data 📊 - An efficient method to train your SLM on this curated data 🔄 3️⃣ What makes it hard? 🤔 Data, of course! While numerous solutions exist to "bake the data into the model," we lack enterprise-grade approaches for converting enterprise data into training data or generating skills data. Sure, you could turn to data vendors, but that route is typically extremely expensive and often struggles to scale as your enterprise data grows. 📈 4️⃣ What tools are available for LLM customization today? ⚡ This is exactly where InstructLab shines! ✨ Unlike other customization offerings, IL provides both tools in one comprehensive open-source package, backed by the expertise of RedHat and IBM. IL features some of the most advanced synthetic data generation methods, enabling you to: - Convert your enterprise data into high-quality training data - Synthesize skills data using open-source LLMs - Leverage novel large-scale training methods to optimize your small model The results? 🎯 On numerous enterprise tasks, including RAG implementations, IL-tuned 7B/8B models consistently outperform proprietary models at a fraction of the cost! 🤝 Let me know if this was helpful, I'll be happy to create a comprehensive long(-er) form post sharing concrete results from real-world implementations. #GenAI #MachineLearning #EnterpriseAI #Innovation #TechTrends #InstructLab #SmallLanguageModels

51

#DayThree Jennifer Doudna, the biochemist renowned for inventing CRISPR, has not explicitly expressed her opinion on Artificial Intelligence (AI). However, she acknowledges the ethical concerns surrounding innovations like CRISPR and AI. In a discussion with Fei-Fei Li, who heads the Human-Centered Artificial Intelligence endeavor, Doudna emphasized the need to consider societal impacts and ethical implications of scientific work. While her focus remains on genome editing, she recognizes that AI will likely have a significant impact, particularly in enabling greater efficiency. #AI #Copilot #MH #captainfuture

2

Thank Saint Louis University School of Science and Engineering for sharing our work! This dual recognition, being named a finalist for both #GordonBell (general track) and #GordonBell for #Climate Modelling, highlights the critical role of the #PaRSEC task-based runtime system, co-developed with University of Tennessee, Knoxville and a key focus of my research, in advancing complex scientific applications. 𝐄𝐱𝐚𝐬𝐜𝐚𝐥𝐞 𝐆𝐖𝐀𝐒: a finalist for #GordonBell, in partnership with KAUST (King Abdullah University of Science and Technology), Massachusetts Institute of Technology, and NVIDIA. By implementing mixed-precision computations and relying on PaRSEC, Exascale GWAS significantly enhances Kernel Ridge Regression (KRR), a critical technique for analyzing genetic data, achieving a performance improvement of five orders of magnitude over the state-of-the-art CPU-only REGENIE, the current leading software for similar tasks. More details: https://lnkd.in/gEUKwYd6. 𝐄𝐱𝐚𝐬𝐜𝐚𝐥𝐞 𝐂𝐥𝐢𝐦𝐚𝐭𝐞 𝐄𝐦𝐮𝐥𝐚𝐭𝐨𝐫: a finalist for #GordonBell for #Climate Modelling, in partnership with KAUST (King Abdullah University of Science and Technology), NSF NCAR - The National Center for Atmospheric Research, University of Notre Dame, and NVIDIA. This exascale climate emulator is to address the growing computational and storage demands of high-resolution Earth System Models. By leveraging advanced numerical methods with the PaRSEC dynamic runtime system, the resulting pipeline, stochastically models spatio-temporal variations in climate data, boosts computational efficiency, greatly enhances the accuracy of climate emulators, and provides critical insights for understanding climate change and extreme weather patterns. More details: https://lnkd.in/gqs_QgHr. Many thanks to our 𝐄𝐱𝐚𝐬𝐜𝐚𝐥𝐞 𝐆𝐖𝐀𝐒 team (Hatem Ltaief, Rabab Alomairy, Ph.D., Jie Ren, Lotfi SLIM, Thorsten Kurth, Benedikt Dorschner, Salim Bougouffa, Rached Abdelkhalek, and David Keyes), our 𝐄𝐱𝐚𝐬𝐜𝐚𝐥𝐞 𝐂𝐥𝐢𝐦𝐚𝐭𝐞 𝐄𝐦𝐮𝐥𝐚𝐭𝐨𝐫 team (Sameh Abdulah, Allison Baker, George Bosilca, Stefano Castruccio, Marc Genton, David Keyes, Zubair Khalid, Hatem Ltaief, Yan Song, Gera Stenchikov, Ying Sun), and the 𝐏𝐚𝐑𝐒𝐄𝐂 team (George Bosilca, Thomas Herault, Aurelien Bouteiller, and Joseph Schuchart)!

54

10 Comments

Gartner is thrilled to unveil the 2025 Top 10 Strategic Technology Trends ✅ For CIOs and senior IT leaders, true leadership is about anticipating and preparing for the future beyond the immediate horizon. I have had the pleasure of leading the research on Hybrid Computing, which is a major trend for 2025, considering the emergence of compute mechanisms with neuromorphic, and quantum. We plan on expanding further guidance on it in the coming months on how enterprises need to plan towards leveraging hybrid computing environment and manage workloads across compute paradigms. Gain a competitive advantage by adopting these trends: https://gtnr.it/3YhS0NQ Gartner for IT | #GartnerSYM #TechTrends #CIO #Technology #AI #FutureofComputing

32

The entire suite of Granite 3.0 models and the updated time series models are available for download on HuggingFace under the permissive Apache 2.0 license. #IBM #opensource #AI

16

#IBM's newly-extended #roadmap shows how we aim to make #quantum_computers more useful in the coming years, as we work toward #quantum-centric #supercomputing. https://lnkd.in/g9s_AAxz ​

1

🚀 Excited to share a new blog post on Amazon Science: "How the Lean Language Brings Math to Coding and Coding to Math"! This post is an adaptation of the keynote I delivered at the International Conference on Computer Aided Verification (CAV) in July 2024. It explores how the Lean language is bridging the gap between formal mathematics and software development, offering new possibilities for both fields. https://lnkd.in/grzRHDHg

158

5 Comments

GLObfus: An Innovative Method of Data Obfuscation to Protect Numerical Data in Public Cloud Storage   D. I. George Amalarethinam a and Lalu P. George b*   DOI: https://lnkd.in/g3Vt5jE6 

8

Peter Lee, President Microsoft Research, is one of the top thinkers on AI within Microsoft, someone whose views are worth deeply reflecting upon. I met him last year during his visit to Microsoft Research India, attended his talk on AI for Heathcare, and found him to be one of the most humble, thoughtful and intelligent human beings. The top 3 trends Peter has shared in this article are: 1: Developing AI systems for science that can learn and understand the “languages” of nature, such as the structures of proteins, molecules, and materials, and combining these with general language models to create powerful multimodal AI technologies. 2: Advancing the field of autonomous AI agents that can plan and execute complex tasks, collaborate with other AI agents, and learn from their actions. 3. Managing AI infrastructure, optimizing hardware architectures, phasing out old systems, and preparing data centers for increasingly powerful AI models and services. Drawing from my first decade of professional career in R&D, I am especially excited about the potential of AI for Science, R&D, and Innovation in ways that were impossible until recently: across chemistry, materials science, life sciences, physics, semiconductors and more... Along with Aseem Datar, I met a few organizations in Mumbai this past week and do think the potential of leveraging AI to accelerate Science, R&D and Innovation in India is huge. #AIforScience #Innovation #ResearchAndDevelopment #India Alok Lall | Venkat Krishnan | Irina Ghose | Puneet Chandok | Venkat Padmanabhan | Sriram Rajamani

69

4 Comments

This is a very exciting milestone for all the great GenAI work taking place at #IBM Research! State-of-the-art performance of the IBM Granite code LLMs and extensive benchmarks results included in the paper: https://lnkd.in/ekkPAVmF

16

Tomorrow, Tuesday, Oct 1, at 6 PM ET, IEEE Maine Joint ComSoc/Computer Society Chapter will join the IEEE Day 2024 celebration by co-hosting a live webinar GENERATIVE AI: FROM CONCEPT TO DEPLOYMENT. As generative AI (GenAI) technologies advance, their impact on industries from healthcare to entertainment is undeniable, with a projected market value exceeding $110 billion by 2030. Gaining an understanding of GenAI solutions is crucial, not just for tech professionals but for anyone looking to stay relevant in an increasingly AI-driven world. We will walk through the core aspects of building GenAI solutions, covering foundational AI models, the art of prompt engineering, and the technical considerations for selecting Large Language Models (LLMs). Engaging in hands-on demonstrations will bring the technology to life. We will address significant challenges and risks, showcasing the dual-edged nature of these technologies. Ethical considerations will be at the forefront, emphasizing the necessity of responsible innovation. Through real-world examples, attendees will gain a clear understanding of GenAI's potential and pitfalls, equipped to navigate and contribute responsibly to this rapidly evolving field. The webinar will be led by our speaker, Mrinal Karvir, Senior AI Software Engineering Manager at Intel Corporation, and followed by Q&A. Please register here:

10