Yunan Jiang

Getting started with AI can seem daunting, especially when contemplating its safety against data privacy. But past that fear lies a fix for itself: AI tools can be harnessed to overcome cybersecurity hurdles. Join us for an insightful discussion at the intersection of security, building resilience, and emerging technologies like AI, IoT, and others such as operational technology, and products intersecting with information technology and engineering. Get inspired, learn something new, and connect with peers, partners, and Googlers!

Getting started with AI can seem daunting, especially when contemplating its safety against data privacy. But past that fear lies a fix for itself: AI tools can be harnessed to overcome cybersecurity hurdles. Join us for an insightful discussion at the intersection of security, building resilience, and emerging technologies like AI, IoT, and others such as operational technology, and products intersecting with information technology and engineering. Get inspired, learn something new, and connect with peers, partners, and Googlers!

Getting started with AI can seem daunting, especially when contemplating its safety against data privacy. But past that fear lies a fix for itself: AI tools can be harnessed to overcome cybersecurity hurdles. Join us for an insightful discussion at the intersection of security, building resilience, and emerging technologies like AI, IoT, and others such as operational technology, and products intersecting with information technology and engineering. Get inspired, learn something new, and connect with peers, partners, and Googlers!

Lots of interesting perspectives in this discussion. A few that stood out to me: Huang doesn't believe in 1 on 1s for his direct staff of 60. His thought is the focus should be on transparency. If someone makes a mistake, everyone should learn from it. Everyone should know what the goals and challenges are all at once. But he also said this does not "scale down" because junior people still need supervision. 60 directs is a lot of people, but it also eliminates 7 layers of management in his view. Conceptually this sounds the same as Ray Dalio's transparency approach at Bridgewater Associates. His rejection of 1 on 1s for his staff suggests a very high level of trust within his executive ranks. That must have developed over time and according to his management style. As he puts it: "You can achieve operational excellence through process, but craft can only be achieved with tenure". Extraordinary things require craft. Other insights: Attends meetings all day but not regular meetings. E-staff are pinch hitters. Problems, idea, brainstorming or creation meetings only. "0-billion dollar markets." If you try to do something very difficult and create something that has never been done before, the effort provides huge opportunities. More fun to create then take share. Also, no one will notice your failures if you are the only one doing something new. If you are not using AI, you will be replaced by someone who is or by a company that is using it. (It might sound like he's talking his own book, but I find AI to provide significant productivity boosts. I agree with his assessment.) Huang talks about the need for an aesthetic in high quality work. Collison agrees. Other examples of people I can think of who said this are Steve Jobs and Jim Simons. I think it is more than coincidence when extremely accomplished individuals have similar principles.

Explore the trade-offs between throughput and latency in #deeplearning models with this new self-paced course. Learn how to measure and optimize performance for on-prem and #cloud deployments. This course is free for a limited time. Enroll now: https://nvda.ws/3O5ui2R #NVIDIATraining

Learn about NVIDIA TAO 5.5: A powerful training framework with new foundation models, multi-modal sensor fusion, and advanced features like text prompt-based auto-labeling and knowledge distillation. Read more. https://nvda.ws/3XjvXY2 #NVIDIAMetropolis

Boost your #LLM inference performance with NVIDIA GH200. Learn how this NVIDIA superchip minimizes trade-offs between user interactivity and system throughput, improving TTFT by up to 2x on the Llama 3 70B model. Read here:

Here’s how @Qualcomm’s new laptop chips really stack up to @Apple, @Intel, and @AMD⁉️🤔 We tested every Snapdragon X chip against the @Intel Core Ultra, AMD Ryzen 8000, and Apple M3. @Microsoft's 12-year-long attempt at making Windows on Arm happen is starting to pay off. Qualcomm's Snapdragon X Elite and X Plus chips are turning Windows on Arm into a viable platform. While the chips don't outright beat Apple's M3 chip in every benchmark, this is the closest Microsoft has been able to compete so far with MacBooks in price, performance, and battery life.