NVIDIA CEO Jensen Huang is about to take the stage for a highly anticipated speech in Taipei, kicking off Asia's biggest tech summit. Asian stocks are climbing to a record as investors double down on the AI trade that has powered equities to all-time highs, despite geopolitical and economic risks. I'm Paul Allen and this is 'Insight.' We have got oil rising after the U.S. and Iran trade messages pushing for changes to a draft deal that can extend the cease-fire and reopen the Strait of Hormuz. Meanwhile, Israel expands its ground assault in Lebanon. Pete Hegseth avoids mentioning Taiwan during a speech in Singapore, which has been seen as a softer stance towards China. Defense chiefs from Malaysia, the Philippines, and Germany are at that gathering.
As mentioned, we are awaiting the appearance on that stage very shortly of NVIDIA CEO Jensen Huang. In fact, he is about to make some remarks. He is speaking about GPU technology. That is video, actually. That is not the Taipei conference, but he will still be taking the stage shortly. He's a local hero in Taiwan, born in the southern city. We are expecting him to talk about the latest products, the role of NVIDIA in Taiwan, and plans announced next week to invest $150 billion a year in Taiwan. Computex is a huge event, 1,500 exhibitors in Taipei from 33 countries. Let's get more on what markets are expecting from NVIDIA CEO Jensen Huang's keynote speech. Minmin Low joins us. I gave a bit of a breakdown of what we might hear about. Did I miss anything?
If you look at the market reaction, a lot of anticipation for Computex. The biggest market movers on the TAIEX can take a hint of what is moving from what happened over the weekend and what is dubbed as the trillion-dollar banquet that Jensen Huang hosted in a meeting with his suppliers. TSMC CEO sitting next to him. You have Asys and MediaTek as well. These are the companies that are moving today on the Taiwan Stock Exchange. You mentioned new product offerings. We will be watching for any comments on the AI computing platform, the new CPU that the company is developing. You mentioned the investment into Taiwan. The company is looking at building out a Taiwan headquarters that would be operational by 2030, so any progress on that that can bring the company closer to its main supplier of TSMC would be of interest as well. It was previously reported NVIDIA was looking at developing an ARM-based PC chip that could pose a challenge to maybe Intel. This is something to look out for. So many exhibitors that will be present there. We will be watching for any new offerings from the other companies that are attending as well, from the giving developers, hardware and software makers. A lot of concern around the challenges that the industry is facing, particularly when it comes to the supply crunch that is hitting not just memory chipmakers but now extending into other critical components of the AI data center buildout. Adjacent to the entire chips industry, we are looking at PC makers. Lenovo, Dell. Goldman Sachs raising the price target for Lenovo. That is a company at the forefront of the drive to make AI PCs. These are the companies that will be gathering in Taiwan to discuss the nuts and bolts of the AI development in the post-ChatGPT world. A lot to look out for ahead of Jensen Huang's speech.
We are expecting him to take the stage soon, but as we await that, you mentioned some of the challenges that the industry is facing in terms of supply chain crunches and geopolitical uncertainty, which is worth mentioning considering this event takes place in Taiwan. The markets seem to be looking through a lot of that, doesn't it?
Yeah, we are seeing this run-up in prices, especially when it comes to the tech names, especially the KOSPI, the TAIEX, which has really overtaken India as among some of the biggest stock markets in the world. You mentioned geopolitical challenges. One of the things that can come in the discussion, but perhaps not so explicitly, but still a focus, is China's ascendance in the AI race, particularly with Huawei closing the gap with TSMC. What kind of challenge will this pose to the big tech-heavy Taiwan and the world? Some of the rivals of TSMC as well. Jensen Huang running onto the stage there as he waves to the members of the audience. I was talking about the rivals of NVIDIA. The likes of Qualcomm, which struck a deal with ByteDance. What these companies would announce as they attend Computex will be in focus as well.
All right, Minmin Low there. Let's take a listen to Jensen Huang on the stage about to make that keynote address.
Superstars. Look how adorable they are. The superstars of Taiwan. There are so many of you here today. We are broadcasting this right now to 70 other watch parties across Taiwan. 70 different conferences are going at the same time. Everybody's watching this. We have so much to tell you. We are broadcasting this right now to 70 other watch parties across Taiwan. 70 different conferences are going at the same time. Everybody is watching this keynote. We have so much to tell you and I have so many partners to thank. It is incredible how large our ecosystem in Taiwan has become. Most of the time when people think about ecosystem, they think about our software stack. They think about the developer ecosystem above the computing systems that NVIDIA builds, but NVIDIA's ecosystem spans all the way upstream to all of our supply chain here in Taiwan where it all begins, and downstream all the way to data centers and eventually to end-users. Today, we will talk about almost all of the ecosystem. There are so many people to thank. I love my ecosystem here. There are so many companies here and some of my favorite ecosystem partners.
Taiwan's rich ecosystem. The richest ecosystem, the world's best supply ecosystem. Unbelievable. Thank you all for being here. This year, our businesses together are growing incredibly. Somebody told me last night that the annual GDP of Taiwan is going to grow almost 10%. Unbelievable. Well, we have a lot to talk about, let's get going. Two years ago when I was here, I started to talk to you about how AI has moved from generative AI and the other waves of AI that are coming. The next wave of AI was agentic AI. And today, we can say that agentic AI has arrived. That useful AI has arrived. What does this mean? This is GitHub. One of the first applications of agentic AI is software coding. One of the most valuable professions. Incredibly large ecosystem. 30 million, 40 million professional software developers. Probably another couple hundred who are students and enthusiasts and so forth. 30, 40 million software developers in the world, code for a living. This represents most of them. This is GitHub. The pull request is when they download software, they modify it. And commit is when they push it back up.
If you can look at this, in 2023, the number of commits was 300 million. 2024, 400 million. 2025, 500 million commits. In the first few months of 2026, it has nearly tripled. What does that mean? 30 million software developers representing about $3 trillion worth of GDP, producing -- that is what they are paid. $3 trillion worth of salaries per year which is generating economic growth for the rest of the industries. Say $100 trillion of the world's interest rates is generated by $3 trillion worth of salary. That $3 trillion worth of salary is now producing nearly three times as much output. It's effectively a $9 trillion productivity from $3 trillion of salaries. Does that make any sense? The difference is absolutely extraordinary. This is the promise of AI. The number of software engineers is increasing. People talk about AI reducing jobs. Complete nonsense. It is causing more software engineers to be hired and the reason for that is simple. If you can hire a software engineer, you can generate $9 trillion worth of productive work, why wouldn't you want to hire more software engineers? If that line was flat, then obviously people would hire fewer software engineers but because the output is so incredible, people want to hire more software engineers. This will show up in our economy somehow soon. The first thing is, useful AI has arrived. What does that mean from the industry perspective? That means that tokens are now in extraordinary demand. If you can do this, you'll want to produce more of it. Because tokens are profitable units, tokens are now profitable units of revenues. Because it is now profitable, the AI companies want to build a lot more tokens, generate a lot more tokens, build more AI factories. Which is the reason why compute demand in Taiwan has skyrocketed. It is precisely the reason why all of you are so busy and your businesses are doing so well. That looks like some of your stock price.
The compute pattern has changed. Everything has changed. The first idea is that useful AI has arrived. AI is now a profit generator, a GDP generator. Behind it is a whole new kind of computing pattern, not just a large language model but an agent. Today, almost everything we are going to talk about is going to be based on this. Let me take a quick moment and show you what I'm talking about. Inside -- this is an agent. An agent application. In the old days, this would be application. This would be code. And this would be operating system. Application, code running inside an operating system. Today, it is an agent which consists of a large language model, many sitting inside a harness and that harness helps it orchestrate it to do productive work. This is the input. When that input comes, it has to understand, observe, reason, act, use tools. That tool could be a spreadsheet, web browser, data processing engine, database engine. This is orchestrated -- this routing of information. Every single time it touches either processing the context, understanding what is happening, reasoning about what to do, coming up with a plan that it acts on. That orchestration path is orchestrated by some software. So, this is fundamentally an agent. It deals with short-term memory called working memory. Long-term memory. The memory management system is incredibly important. This entire system is called an agent. The large language model is used to do the thinking and the harness connects everything together, just like an operating system. This is the new computing model and this is what an agent, it can do incredible things. This is the big breakthrough. The convergence of large language models that are now able to do a really good job thinking, reasoning, planning, using tools, and the fact that we have now these harnesses that manage memory, the orchestration, uses tools, we can now do amazing things. Let me give you an example. This is a prompt. This is the prompt. This is the code that is generated. This comes out. This is the input. This is the input. And that is the output. What do you guys think? It's pretty amazing, right?
Use Claude Code here but Codex is an incredible job as well. Here is another example. This is the input. NVIDIA logo and then scatter and repeat, right? So you saw that. That was the prompt. Here is the next one. I lost my remote control battery clip. It looks like this. Create a CAD file that uses the tool, ready for printing to create a new one. Make sense? This is now the new computing pattern. We used to launch an application, click and type. We now replaced that with explaining to the AI what we want, our intent, and the AI generates the code or uses tools to produce the necessary output. This is how computers are going to work in the future. This is agentic AI. For two years, we have been building towards this, and now, it has arrived. One of the big breakthroughs of course is tool use. A lot of people have said AI is coming. Agentic AI is coming and therefore all the software companies are going to go out of business. I said, it's exactly the opposite. Because there's going to be so many agents, the world is no longer limited by the number of people. Therefore, those agents are going to use more tools than ever. This is actually an incredible time to be a software company. But the software has to be presented to the agent in a way that the agent can use it. This is a big breakthrough and what we have done, as you know, why NVIDIA's treasure is all of our libraries. This is NVIDIA's treasure. Today, we are able to now present these libraries to agents who can use it much more effectively than even humans so this is a wonderful time. Let's take a look.
20 years ago, we built a single architecture for accelerated computing. We reinvented computing. 1,000 libraries helped developers make breakthroughs in every field of science and engineering. They are tools for agents. Computational lithography. Decision optimization. Direct sparse solvers. cuDNN for deep research across structured and unstructured documents. Ariel for AI. Warp for physics. Genomics. At their foundation, our algorithms. And they are beautiful.
Just what we were waiting for, Jensen Huang to resume speaking. Quickly recap some of the points he has been making. He started off by hailing Taiwan's supply chain and tech ecosystem but then he was talking up agentic AI, saying useful AI has arrived as well. He's saying we were seeing the enormous growth in productivity explode with AI, causing more software engineers to be hired so AI is having a multiplier effect on some of the things that software engineers can do. Jensen is about to resume speaking again so let's listen in.
A round of applause for math. Math is beautiful. The computing pattern of software is going to change. In fact, let's come back to this. This is the agent. It is the ultimate disaggregated and distributed computing model. So many different computers are going to be activated in order to process this agent. The agent consists of model, harness, tools and skills, and a runtime. All of that is running at different places in a data center. You can think of the model as the brain. The harness as the body. The toolset uses working in a runtime -- think of it as a workshop so this is a person, a worker working with tools and a workshop. Of course, this is being done at extraordinarily large scales and each one of those steps are running in a different part of the computer. And you can see the large language model is thinking, context processing, observing, understanding the environment, reasoning, coming up with a plan, and acting on the plan. Every single time that happens, an entire rack is activated. It is thinking with the large language model. Whenever it uses a tool, a CPU is used. That tool could be a compiler, Python. It could be JavaScript or it could be accelerated computing. Today's agents are relatively simple users of tools. Tomorrow, they are going to be very sophisticated users of tools which is the reason why the libraries I showed you are going to be incredibly popular with agents. They solve some of the most important problems the world knows. And all of our libraries are now going to come with skills that the AI could learn how to use. So the library, some skills, basically a manual, the AI reads it and goes that's how you use it. The ability to use these libraries by agents are going to be incredible so the tools run on CPUs and chips used in models. The security harness runs on CPUs and a security processor called a DPU, BlueField. The orchestration of all of this runs on a CPU and this is the entire harness and the CPU is orchestrating all the work. One of the hardest parts is memory. You could just imagine. What to remember, compaction, not just compression, but how to retrieve. Do you retrieve structured data? Unstructured data? What is the ontology, the relationship of all of the different data to itself? The entire processing is incredibly complicated. The memory system, the memory system of AI is going to cause the storage system to be completely revolutionized. As you can see, every aspect of this computing model, this computing pattern, this new application called an agent is fundamentally different. Disaggregated, distributed, heterogeneous computing problem is precisely the reason we built our next generation. Vera Rubin is not one chip. It's not a GPU only. It starts with a GPU but Vera Rubin is incredible. This entire thing is Vera Rubin. From end to end, it is GPUs. It is orchestrated -- something to tell you more about. The storage systems, revolutionary. The security processor that is inside so that everything is encrypted at rest, in motion as well as in use. Everything across this is secure because the AI model is so precious and this is the reason why this entire system obeys confidential computing. Each one of these systems would be a complete revolution in itself. The most ambitious endeavor in the history of our country. The whole company worked on it across all 40,000 engineers and all of you participated in the creation of the entire system.
NVIDIA used to be a GPU company and over the years we have evolved to become a systems company and you are looking here now for the most complex system, the most complex system ever designed but ultimately, our customers, our partners don't want to buy computers. They want to build AI factories which is the reason why NVIDIA has really started to transform ourselves yet again. You can see so much of our technology is now at the entire infrastructure scale. Our partners are at infrastructure scale. Power generators, cooling systems, the grid providers. So many industrial companies are now part of our ecosystem because ultimately, we are trying to build an entire stack just like GPUs, just like when we are building Grace Blackwell and just like now, we are building a full stack system so that our customers could build amazing AI infrastructure.
We will just recap some of those remarks from Jensen Huang's speech so far. He spent quite a bit of time explaining agentic AI tools, how it works, also talking about how every time somebody uses AI, technology activates. Ian King was talking recently about this kind of approach and he does see his speeches as lectures to educate people about AI and NVIDIA and certainly the remarks so far do tend to be following that tone and he did also hail Vera Rubin just before he left. This video describing it as a miracle and also praising those people involved. He had interesting remarks a little bit earlier about the impact of AI on software engineers as well. He doesn't see this as a threat to software companies but rather as a multiplier in many ways. They can use tools more than ever. It's an incredible time to be a software company but software does need to be presented to an agent in a way that that agent can use it. So these are some of the remarks we have had from Jensen Huang so far and this is the keynote address of course at the conference in Taipei. There are 1,500 exhibitors present from 33 countries and they are all hanging on every word from Jensen Huang but he appears to be showing these videos at some point to catch his breath but just to recap what he's saying, he says useful AI has arrived and he was also describing situations where people soon can now use prompts which can result in creation of a CAD file that can then be given to a 3D printer to print difficult to replicate replacement parts for equipment and other pieces of equipment that might be broken so talking about some of the practical implications of some of these AI tools that NVIDIA is working on as well. So talking about the agent, he gave us quite a long breakdown of how this works, talking about a model, harness, tools, skills, and runtime and these are the factors that make up an agent. He says this happens at different places in a data center simultaneously and he described the model as the brain and the harness as the body and the tools in the workshop, really playing into that idea. He treats it as an opportunity to educate the public about AI and NVIDIA more broadly. Jensen Huang returning to the stage now. Let's listen.
