On the evening of June 2nd, Nvidia's Jensen Huang delivered the keynote speech for the 2024 Computex at the National Taiwan University Sports Center. Following last year's theme of "This is the iPhone moment for AI," this year Huang set the tone with "Accelerate Everything." Despite the heavy rain in Taipei, over 6,500 attendees, including leaders from Taiwan's tech and financial industries, packed the NTU Sports Center.
Huang stated, "Generative AI is reshaping industries, bringing new opportunities for innovation and growth."
"We are at the cusp of a major shift in computing today," Huang, dressed in his signature black leather jacket, told the audience. "Artificial intelligence and accelerated computing will redefine the future."
The National Taiwan University Sports Center gathered over 6,500 industry leaders, media representatives, entrepreneurs, gamers, creators, and AI researchers.。
"The future of computing is accelerated," said Jensen Huang. "Through our innovations in artificial intelligence and accelerated computing, we are pushing the boundaries of what’s possible and driving the next wave of technological advancements."
Here are the highlights of the keynote speech:
With a "One Year" Cadence
Jensen Huang revealed a new semiconductor development roadmap, which will be released on an annual cycle.
The newly unveiled Rubin platform will succeed the upcoming Blackwell platform, featuring new GPUs, a new Arm-based CPU called Vera, and advanced networking technologies, including NVLink 6, CX9 SuperNIC, and the X1600 converged InfiniBand/Ethernet switch.
Huang stated, "Our company operates on a one-year cadence. Our fundamental philosophy is very simple: build at the scale of an entire data center, break it down, and deliver products to you on an annual basis, pushing everything to the technological limit."
The More You Buy, The More You Save!
"Accelerated computing is sustainable computing," described Jensen Huang. The combination of GPUs and CPUs can achieve up to 100 times the acceleration while only increasing power consumption by threefold.
Jensen Huang stated that as accelerated computing becomes more powerful, the processing speed of CPU+GPU combinations will increase by 100 times, while power consumption will only increase threefold. (Source: Screenshot from Nvidia YouTube)
"The more you buy, the more you save," Jensen Huang pointed out, emphasizing that this approach will lead to significant savings in both cost and energy.
Partnering with the Industry to Build AI Factories and Drive a New Industrial Revolution
This speech was filled with mentions of Taiwan.
Jensen Huang pointed out that the world's leading IT manufacturers, particularly those from Taiwan, have already adopted Nvidia's GPU and networking solutions. Companies such as ASUS, Gigabyte, Inventec, Pegatron, Wistron, Wiwynn, and international giants like Supermicro are creating cloud, on-premises, and edge AI systems.
Nvidia NIM Transforms Millions into Generative AI Developers
With Nvidia NIM, 28 million developers worldwide can easily create generative AI applications. NIM can be deployed in the cloud, data centers, or workstations.
NIM also enables enterprises to maximize their infrastructure investments. For example, running Meta Llama 3-8B on NIM can generate three times more AI tokens on accelerated infrastructure compared to not using NIM.
Currently, nearly 200 technology partners are integrating NIM into their platforms to accelerate generative AI deployments for specific applications, such as assistants, code assistants, and digital human avatars. "Today we just released the fully optimized Llama 3 on Hugging Face; you can try it out and even take it with you," said Jensen Huang. "You can run it in the cloud, any cloud, download the container, put it in your data center, and offer it to your customers."
AI PC Brings AI Assistant into Your Life
The AI PC steals the spotlight at this year's Computex!
Jensen Huang pointed out that AI PCs equipped with Nvidia RTX will completely transform consumer experiences.
Nvidia, known for its powerful graphics cards, considers RTX one of its most significant computer graphics technologies. Through AI, RTX can accomplish 3D design and simulations faster and with superior visual effects than ever before.
Nvidia also announced Project G-Assist—a RTX-powered AI assistant technology that provides contextual awareness for PC gaming and applications.
IT industry leaders are leveraging Nvidia's AI robotics technology.
Taiwanese manufacturers are utilizing Nvidia's robotics technology to transform their factories.
Jensen Huang demonstrated how Foxconn is using Nvidia Omniverse, Isaac, and Metropolis to create digital twins, combining visual AI and robotics development tools to enhance the efficiency of robot factories.
AI robots will be ubiquitous.
"The next wave of AI is physical AI, AI that understands the laws of physics, AI that works alongside us," emphasized Jensen Huang on the importance of robotics technology and AI in future developments.
The Nvidia Isaac platform provides developers with powerful toolkits for building AI robots, including AMRs, industrial robotic arms, and humanoid robots, all driven by supercomputers and AI models such as Jetson Orin and Thor.
"Robotics technology is here, physical AI is here, this is not science fiction, and it's already being used across Taiwan. It's really very exciting!" added Jensen Huang.
Jensen Huang shared that many of Taiwan's tech giants are applying Nvidia's robotics technology and expressed that AI humanoid robots will become increasingly commonplace in the future.
Global tech giants are integrating Nvidia's autonomous robotics technology into their factories, utilizing simulations in Omniverse to test and validate the next wave of real-world AI.
"All factories will be robotized. Factories will command and deploy robots, and these robots will manufacture robotized products," explained Jensen Huang.
Jensen Huang emphasized the role of Nvidia Isaac in enhancing factory and warehouse efficiency, with global leaders such as BYD, Siemens, Teradyne Robotics, and Intrinsic adopting its advanced AI models.
Nvidia AI Enterprise on the IGX platform, in collaboration with partners like ADLINK, Advantech, and ONYX, offers edge AI solutions that meet strict regulatory standards, crucial for industries like healthcare technology.
Taiwan is an unsung hero, a cornerstone of the world.
At the end of the speech, a video titled "Thank You, Taiwan" ignited the audience present at the event and those watching the livestream.
"Taiwan, you've helped us realize a vision. Along this long journey, countless partners have supported us, from accelerated computing, computer graphics, scientific research to the pinnacle of artificial intelligence. Every chip, every computer, tells the story of your hard work and pursuit of perfection.
Taiwan is an unsung hero, yet a pillar of the world.
Together, we've propelled the industry's revival, toiled tirelessly, allowing us to step into the realm of cutting-edge computing. Among these, the computers we produce will transform into new types of factories, manufacturing invaluable artificial intelligence. Dear partners, you are the backbone of AI industry innovation, steadfast as a rock in the face of any storm.
Your excellence in technology, diligence, and spirit of unity is your superpower. Without you, our vision would be nothing but an unattainable dream."
In the resounding conclusion of this captivating keynote speech, Jensen Huang once again paid tribute to Taiwan and Nvidia's numerous partners: "Thank you, I love you!"
Jensen Huang's COMPUTEX keynote speech, all 43 enterprise names displayed on the big screen at once.
Comments