Source: Jinshi Data
As NVIDIA's Market Cap approaches the "top position" once again, Huang Renxun released several heavyweight new products at the "Technology Spring Festival." Will the stock price welcome a new round of explosion?
At the 2025 CES (International Consumer Electronics Show), known as the "Spring Festival Gala of Technology," the CEO of the world's second-largest Market Cap company, Huang Renxun, delivered a keynote speech. $NVIDIA (NVDA.US)$ He released a series of new Hardware, new Software, and new services, including consumer-level GPUs based on the Blackwell architecture and the foundational model NVIDIA Cosmos, aimed at understanding the real world, and announced that Toyota Motor and NVIDIA will collaborate to develop the next generation of autonomous vehicles.
Huang Renxun stated that AI technology is playing a dominant role in improving the training of robots and autonomous vehicles, as well as in new gaming chips, which marks the potential for the company's Business expansion. He also mentioned that in the upcoming years, the amount of data generated by humanity will exceed all data produced from ancient times to the present, and this data is becoming multimodal, including video, images, and sound.
Consumer-grade GPUs based on the Blackwell architecture.
Huang Renxun announced the consumer-grade GPUs based on the Blackwell architecture, with the prices of the GeForce RTX 5090, RTX 5080, 5070 Ti, and 5070 being $1999, $999, $749, and $549 respectively, set to be launched later this month. It is reported that the GeForce RTX 50 series has 92 billion transistors, 4000 TOPS, and four AI processing units, with performance three times that of the previous generation Ada. The performance of the RTX 5090 will be twice that of the 4090, showcasing a significant leap in the capabilities of NVIDIA's 5070 to 5090 GPU series.
The latest gaming chips utilize NVIDIA's Blackwell AI technology, providing video games with cinematic graphics, especially in the area known as 'shaders,' which can make images like ceramic teapots appear more realistic by adding imperfections and fingerprints on the surface.
Laptops equipped with GeForce RTX 5090, RTX 5080, 5070 Ti, and 5070 are priced at $2899, $2199, $1599, and $1299 respectively, and will be launched in March.
Jensen Huang stated that machine learning fundamentally changes how applications are built and computed, expanding possibilities beyond traditional methods, while GeForce GPUs play a key role in advancing AI for public use.
Cosmos base model
NVIDIA also launched the Cosmos base model aimed at understanding the real world, which can accelerate the training of robots and other physical AI systems. The Cosmos model is the first base model in the world, based on 20 million hours of video training, focusing on physical dynamics and human interaction. It can accept prompts in the form of text, images, or videos to generate virtual world states, serving as unique video outputs for the specific needs of autonomous driving and robotics applications.
NVIDIA Cosmos can generate realistic videos at a far lower cost than traditional data for training robots and autonomous vehicles. By creating 'synthetic' training data, these models can help robots and cars understand the physical world, similar to how large language models assist chatbots in generating replies in natural language. Users only need to provide a text description to generate videos that adhere to the laws of physics. Developers can leverage Cosmos to generate AI feedback for reinforcement learning, thereby improving strategy models and testing performance in different scenarios.
This method will be significantly cheaper than the current cost of collecting data. Currently, to train autonomous vehicles, companies need fleets to patrol the streets to collect data, while humanoid robots are typically trained by having real people repeatedly perform tasks.
However, Jensen Huang warned that the Cosmos model needs more data before reaching its 'ChatGPT moment.' Cosmos will provide 'open licensing,' similar to Meta Platforms' Llama3 language model. Jensen Huang stated, 'We sincerely hope that (Cosmos) can bring the contributions that Llama3 made to enterprise AI to the robotics and industrial AI world.'
Jensen Huang also emphasized that the future of robots depends on creating an effective world model that is not only limited to language understanding but also incorporates physical dynamics such as gravity and friction.
Toyota Motor and NVIDIA will collaborate on developing the next generation of self-driving Autos.
Jensen Huang announced that Toyota Motor and NVIDIA will work together to develop the next generation of self-driving Autos. Toyota will use NVIDIA's automotive chips and operating system.
He stated that Toyota Motor will use its Orin chip and automotive operating system to provide advanced driving assistance features for multiple models, although details about these models were not provided. Toyota, AURORA, and Continental Group have joined NVIDIA's ever-expanding list of partners to jointly launch the next generation of highly automated and self-driving fleets. Domestic car manufacturers collaborating with NVIDIA also include BYD, Li Auto, Xiaomi, and ZEEKR.
Jensen Huang expects that by the fiscal year 2026, revenue from automotive Hardware and Software will increase from this year's 4 billion dollars to 5 billion dollars. He also announced the launch of the next generation automotive processor "Thor," a revolutionary robotic computer designed to handle large amounts of Sensors data. Additionally, NVIDIA's Drive OS has become the first certified programmable AI computing functional safety software that meets the highest automotive safety standard ASIL-D.
"Project Digits" Personal Computer
Jensen Huang also showcased a desktop computer named Project Digits. This computer will be equipped with the same chips found in enterprise Datacenter products, but it comes with a central processing unit from MediaTek. These chips will feature a smaller package, enabling individual Software Development developers to quickly test their AI systems.
The initial Project Digits system is not fully aimed at consumers, as it will be running a Linux-based Nvidia operating system, primarily used by computer programmers rather than ordinary consumers, with a price of $3000.
Will the market's expectations be met?
NVIDIA's stock closed at a record high of $149.43 on Monday, bringing its Market Cap to $3.66 trillion, making it the second largest publicly traded company in the world, second only to $Apple (AAPL.US)$ . Investors hope that Jensen Huang's latest speech can drive the stock price to break through again.
Investors are closely watching any comments regarding the Blackwell chip, as it is seen as NVIDIA's next major growth driver. Despite strong demand, the supply of Blackwell is limited due to production challenges, delaying its launch. Matt Cioppa, portfolio manager at Franklin Templeton Equity Group, said:
"Demand for Blackwell is expected to remain very strong. This could refocus the market on NVIDIA's long-term opportunities."
Investors have reason to be optimistic. Over the past six months, Jensen Huang's comments about chip demand have been driving the stock price up. In October, he described demand for Blackwell as "crazy," and in November stated that demand was "very strong," with shipments expected this quarter.
The stock recorded a monthly decline in December, but it has still risen 171% in 2024, making it one of the biggest drivers of the S&P 500 Index. The stock has already increased by 11% this year. With a Market Cap of $3.66 trillion, it is close to surpassing Apple again to become the largest company by Market Cap. Apple's Market Cap is $3.7 trillion.
Currently, NVIDIA remains Morgan Stanley's preferred stock, with Analysts believing that the chip manufacturer will gain market share this year. In a research report last month, Analysts wrote: "The demand for Blackwell is exceptionally high, but supply is limited. We believe that by mid-year, the focus will still be on Blackwell, which will be a revenue driver in the second half of the year."
However, with the surge in stock prices, the risk associated with NVIDIA Stocks may have increased, raising concerns about valuation. The stock is trading at 19 times estimated earnings, making it one of the most highly valued components in the NASDAQ 100 Index. Emily Roland, co-chief investment strategist at John Hancock Investment, has a positive outlook on NVIDIA and other large Technology stocks, but she expects "more volatility in the future."
编辑/jayden