Julio C. Tapia

🚀 Exciting news! #Cohesity adopts NVIDIA NIM to supercharge retrieval augmented generation apps for enterprises. 🚀 This collaboration enhances #AI capabilities: 1️⃣ Faster deployment of generative AI models 2️⃣ Streamlined development with prebuilt microservices 3️⃣ Accelerated AI innovation for your enterprise

1

Check out this new SIGMOD '24 scientific paper. Our customers love Amazon MemoryDB for its low latency performance, rich flexible API, large scale (up to 500 nodes per cluster), security, availability (4 9s SLA), and AWS's operational excellence. We designed MemoryDB for 11 9s of durability while still providing in-memory performance. My team created a nice technical scientific paper on MemoryDB's architecture. If you would like to learn about what's under the hood, check out this SIGMOD '24 paper. Yacine Taleb, Xingbo Wang, Nan Yan, Allen Samuels, Kevin McGehee, Stefan Mueller, and Jean G. - excellent work publishing this paper! https://lnkd.in/gFXPvci8 #aws #memorydb #inmemory #databases #durability #science #sigmod

192

6 Comments

On #CNBC's #MadMoney, Jim Cramer digs deep with Anirudh Devgan on the importance of Cadence in the development of the #AI era. Hear Anirudh explain how #Cadence’s vision and leadership in computational software are transforming multiple industries and powering the AI era. https://ow.ly/Ow9L50S3PGw

88

2 Comments

I am excited to announce the completion of my Professional Certificate in Machine Learning and Artificial Intelligence from UC Berkeley Executive Education. This program has equipped me with the essential knowledge and skills to excel in the dynamic realms of AI and ML. Offered in partnership with UC Berkeley’s College of Engineering and Haas School of Business, the course delves into key areas such as: - Foundations of Machine Learning and AI: - Introduction to Machine Learning - Fundamentals of Data Analysis - Practical Applications using Python, Jupyter, pandas, Seaborn, Plotly, and GitHub - Machine Learning and AI Techniques: - Decision Trees, Gradient Descent, and Optimization - Clustering, Principal Component Analysis, and Model Selection - Regression Models, Time Series Analysis, and Forecasting - Advanced Topics and Capstone Project: - Deep Neural Networks, Natural Language Processing, and Recommendation Systems - Ensemble Techniques like GBM, XGB, and Random Forest - Capstone Project for real-world problem-solving The interactive group discussions and hands-on coding activities in the program make it a valuable choice for professionals seeking advancement in this high-demand field. Whether you are in IT, data analysis, a recent STEM graduate, or an enthusiast of AI and ML, I highly recommend this program for building a strong foundation and gaining practical experience in leveraging ML/AI tools to solve complex challenges. For those keen on exploring the world of AI and ML, I encourage you to discover more about the course: https://lnkd.in/gr9_4PSS  #ProfessionalCertificate #UCBerkeley

68

12 Comments

OCI is leading the pack in cloud infrastructure performance with its ability to support HPC, Exadata, Autonomous Database, and GPU-powered AI and ML applications. What sets OCI apart is its scalable, low-latency network that supports remote direct memory access (RDMA) for lightning-fast performance. Check out this article for more details on how OCI is building a high-performance network in the public cloud. #OCI #cloudinfrastructure #HPC #AI #machinelearning #Exadata #AutonomousDatabase #GPU #publiccloud #oracle https://lnkd.in/gY4SkF-t

8

We live in an exciting era of pervasive intelligence, where we are seeing innovation and growth at unprecedented speeds. In this blog, I share how Synopsys plays a mission-critical role in enabling silicon and systems companies to meet AI demands through AI-driven EDA, accelerated computing, silicon-proven IP, and multi-die designs. I'm also excited to announce our production-ready multi-die reference flow in partnership with Intel Foundry! Learn about the details at the link below. #Synopsys #61DAC

256

8 Comments

Blackwell 26,596 times more powerful than our 8 Intel MOSIPS? At Computex 2024, Jensen Huang unveiled the roadmap for Blackwell, promising a future of accelerated computing architectures that redefine our expectations. These advancements are set to elevate AI capabilities to unprecedented levels. Our A7Technology MOSIPS system, powered by 8 Intel NUC computers, delivers 752 GFLOPS, with each system offering 94 GFLOPS. This impressive setup, however, pales in comparison to the capabilities of the new Blackwell B200 GPU. The B200 not only doubles the transistor count of the current H100 but also delivers an astounding 20 petaflops of AI performance from a single GPU, compared to the H100's 4 petaflops. The B200's 192GB of HBM3e memory and 8 TB/s bandwidth are equally impressive, showcasing how scaling is driving AI capabilities to new heights. This leap in performance is a significant phase shift, surpassing the limits of Moore's law and pushing the boundaries of what we thought possible with CPU architectures. To put this in perspective, a single Blackwell B200 GPU's 20 petaflops make it approximately 26,596 times more powerful than our current system's 752 GFLOPS, which uses eight of the latest Intel CPUs. These breakthroughs in parallel architectures are not just incremental improvements; they represent a monumental leap in AI capability and power, setting the stage for the next era of technological advancement.

1

Nvidia has surpassed Microsoft to become the world's most valuable company, reaching a record valuation of $3.34 trillion. Daniel Bathurst, Head of Product at Nscale highlights the importance of hashtag #GPU compute and the need for competition in the AI chip market to promote innovation in the following Independent article: “Powerful hardware is the foundation upon which the AI revolution is built” “But, while powerful chips are a key ingredient, the AI ecosystem is more than just hardware. We need a diverse offering at the chip, hardware, and software level to support a thriving, sustainable AI landscape”

10

Extremely delighted to see Jensen launch ACE (Avatar Cloud Engine) platform through the NVIDIA NIM Inferencing Microservices at #COMPUTEX2024. Digital Humans aka Avatars as a channel/medium for Natural Language conversations and interactions is a very interesting product idea that was fun and insightful to collaborate with product/engg teams along with a few select early access Customers and GSI/ISV Partners. The amount of learning I personally went through in the evolution journey of the product was amazing. It feels awesome to see the product line grow and evolve while harnessing the power of LLMs and GenAI technologies from model training to inferencing. At some point in the near future model maturity, stability and predictability will meet human-like experience avatar technologies (A2F, TTS, STT) to deliver human-machine, machine-machine interaction possibilities in the real and virtual world (#metaverse)

18

Amazon Web Services (AWS) is at war and Andy Jassy has a new consigliere Matt Garman. Amazon.com Inc. has strategically positioned Matt Garman as the new CEO of Amazon Web Services (AWS), much like a wartime consigliere in the vein of the iconic film "The Godfather." Garman assumes command in a critical era for AWS, marked by intense competition for cloud and AI supremacy. This period follows significant shifts brought about by the pandemic, a new wave of developers, and rapid advancements in generative AI technology. AWS finds itself in the midst of a fierce battleground, not only maintaining its cloud infrastructure leadership but also vying for dominance in the burgeoning AI sector. The one other wildcard in favor of AWS with Garman (besides being a product czar) is that he's very partner friendly which will be a key part of keeping AWS ahead. Lots of commentary and our post has the news with reaction from the top analysts. SiliconANGLE & theCUBE has the news. Story Link: https://lnkd.in/gDJFC496

49

12 Comments

Easily implement Llama3, SDXL, and other models on Qualcomm Cloud AI 100 using Qualcomm-TensorOpera APIs with dedicated endpoints. Get started here: https://lnkd.in/g2SRDpqs #machinelearning #ai #generativeai

49

The Future of Compute: Scaling Inference with AMD With the latest reports highlighting the incredible power of the AMD MI300X, it’s time to explore this untapped resource. We’ve assembled a panel of AI experts to discuss the emerging ecosystem, early potential, cross-platform benefits, and how you can implement the latest updates in your inference workloads. https://lu.ma/z5zq966l #amd #epyc #instinct

45

2 Comments

One of the best things about being in San Francisco 🌉 is the ability to drop in on great talks. Last week I had a chance to see NVIDIA's CEO Jensen Huang (JH) being interviewed by Stripe's CEO Patrick Collison. Jensen is among the longest-serving founder-CEOs in Silicon Valley at the moment (30+ years). With that comes a special clarity of mind and some hot takes. 🌶👇 🗣 Jensen doesn't believe in 'praise in public and criticism in private'. Instead, he reasons that most information in the organization should not be privy to certain people. By giving feedback in public, you allow everyone to learn. 'What is better than learning from your own mistakes? It's learning from someone else's mistakes!!' 🔥 JH also doesn't like to fire people. He would rather "torture them into greatness". He invests in people and he doesn't like to write off his investments. 👨‍💻 Asked whether he likes his job every day, he answers "no". The perfect job doesn't give you happiness 100% of the time. "No great things were done, that were easy all the way." ... "Even if I come home from a day where I did things I didn't enjoy, it could still be have been a great day." 👩‍👧‍👦 Jensen famously has 60 direct reports. "This allowed us to cut out probably 7 levels of management". His subordinates don't necessarily have such large teams. "It doesn't scale down, less senior people need more guidance". Looking at the numbers, this approach allowed NVidia to reach ~2T market cap with ~28.000 people vs. Microsoft's 3T with ~220.000 people and Apple's 2.6T with 160.000 people. ❔ Jensen answers almost every question in the same way: "Let's break it down, let's reason about it". This is a great response on so many levels. First, it allows him to not be put on the spot. Second, it allows him to teach his workforce how he reasons through a problem so that they can emulate it and create alignment. Third, it gives people a chance to point out flaws in his reasoning. He admits he is not right all the time. ON AI 👨‍🔧 "You will not lose your job to AI, your company will go out of business because another company used AI." "If you're not using AI you're doing it wrong." ⏱ The extent to which AI will enter many processes is largely underestimated. Right now we are focused on language - sounds in a time sequence. "Any time sequence can be tokenized." This will include robotic movement. 🥛 JH also expects that we won't be satisfied with foundation models "You're going to want to tune it into perfection because you care so deeply. Say if it's already at 99% you'll want to get it to say 99.3%" This last piece is also why we created Query Vary (YC W22). AI will never feel truly intelligent until it completely understands your data and context. It needs to be taught, regardless of how powerful the foundation model is. Finally, besides insights, it was remarkable how funny Jensen was, cracking a joke at roughly a 1 per 3 min rate. It was a very entertaining talk. #ai #sf

61

8 Comments