news

tencent technology news according to foreign media reports, nvidia's soaring stock price has pushed it into the top three in the world in terms of market value, but this is actually based on a series of assumptions. first, semiconductor-driven artificial intelligence applications are gradually establishing their position as the new pillar of the modern economy. second, nvidia and its supply chain partners need to demonstrate outstanding capabilities to seamlessly connect and meet the market's explosive growth in demand for high-end computing devices.

nvidia has become the new "chip king", while the "old king" intel is facing many crises and there are even rumors that it will be sold to qualcomm.

in the surge of artificial intelligence, nvidia has firmly established itself as the gold medal "tool supplier" in the "gold rush" with its keen insight and technological innovation. the company's revenue is still soaring, especially the hopper chip series and its highly anticipated successor blackwell, with growing orders.

however, whether nvidia can continue to write its glorious chapter depends on whether it can work with technology giants such as microsoft and google to jointly tap the unlimited commercial potential of artificial intelligence and help these giants transform their huge investments in its high-performance chips into real commercial value.

at the same time, as nvidia's market position becomes increasingly prominent, antitrust regulators are also focusing their attention on it, launching in-depth investigations into whether nvidia uses its dominant market position to set up barriers and prevent customers from switching to other suppliers.

on august 29, nvidia released a disappointing sales forecast, and its stock price suffered its biggest one-day drop in four weeks, with the market reacting strongly. at the same time, the company also officially confirmed the previously circulated news that the design of the blackwell chip is undergoing a key adjustment to optimize the manufacturing process and improve production efficiency.

despite this short-term setback, nvidia's stock price has since rebounded strongly from its august lows. industry analysts predict that nvidia's revenue is expected to double this year, continuing its high growth trend since 2023. barring major unforeseen market fluctuations, nvidia is expected to consolidate its leading position as the world's most valuable chipmaker by the end of the year.

so how did nvidia become the king of chips? and what challenges does it face?

1. which is nvidia’s most popular ai chip?

nvidia's most profitable product on the market is the hopper h100, which is named after grace hopper, a legendary figure in computer science, in order to pay tribute to the latter. as a top-level enhanced version of the gpu (graphics processing unit) in personal computers, the hopper h100 helps video game players get the most realistic visual experience. however, as technology continues to evolve, this star product is about to usher in its successor, the blackwell series, which is named after the famous mathematician david blackwell.

both hopper and blackwell use nvidia's unique chip cluster technology, which transforms multiple chip units into a single efficient whole through high integration. these units can process massive amounts of data in parallel and achieve amazing high-speed computing capabilities. this makes them ideal for the energy-intensive task of training neural networks, which is the key technology for building the latest generation of artificial intelligence products.

nvidia has been at the forefront of artificial intelligence since its founding in 1993, and its forward-looking investment strategy dates back nearly two decades, when nvidia saw that parallel processing capabilities would greatly increase the value of chips in a wide range of applications beyond gaming.

looking ahead, nvidia plans to promote the blackwell series with a diversified strategy, the most eye-catching of which is the gb 200 super chip. this chip cleverly combines two blackwell gpus with a grace cpu designed for high-performance computing, aiming to bring users an unprecedented computing experience and performance leap.

2. why is nvidia’s ai chip so special?

with their powerful learning capabilities, generative ai platforms have shown extraordinary potential in diverse tasks such as translating texts, writing reports, and even synthesizing images. these platforms ingest massive amounts of data and continuously optimize their algorithms, making them perform better and better in complex scenarios such as understanding human language and writing job applications. this process is essentially based on repeated trials and iterations of massive amounts of data. through billions of attempts and deep mining of computing resources, they gradually approach and surpass human-level proficiency.

as a leader in this field, nvidia blackwell has remarkable efficiency in training artificial intelligence. according to official data, its performance is 2.5 times higher than that of the previous generation flagship product hopper. however, behind this outstanding performance is an unprecedented challenge of transistor density - the number of transistors integrated in blackwell is so huge that it has far exceeded the limits of traditional production processes. to this end, nvidia innovatively adopted a dual-chip design and used sophisticated connector technology to seamlessly integrate the two chips into an efficient and coordinated whole.

for customers eager to train ai platforms to perform new tasks, the performance leap provided by the hopper and blackwell series chips is undoubtedly crucial. these high-performance components have become the core driving force for the innovation and application of ai technology, and their strategic value is so high that they have attracted international attention and restrictions, such as export controls imposed by the us government on specific countries.

3. how did nvidia become a leader in ai?

as the king of graphics processing units, nvidia has long been leading the innovation of computer visual experience. gpu is the core component of the computer responsible for generating the images seen on the user's screen, and its technical strength directly determines the visual effect. nvidia's top graphics chips integrate thousands of high-performance processing cores that work in parallel and can handle multiple computing tasks at the same time, modeling complex 3d scenes at an astonishing speed, including fine shadow rendering and realistic physical reflection effects.

entering the 21st century, nvidia engineers showed extraordinary foresight, and they keenly realized that these chips, originally designed for graphics acceleration, actually have a wider range of application potentials due to their powerful parallel computing capabilities. at the same time, researchers in the field of artificial intelligence were pleasantly surprised to find that the intelligent algorithm acceleration they had long pursued could actually be cleverly achieved with the help of these gpus.

since then, the combination of nvidia gpu and artificial intelligence has become the focus of the industry. by continuously optimizing the gpu architecture and software ecosystem, nvidia has not only consolidated its dominant position in the field of graphics processing, but also opened up new markets in the field of artificial intelligence computing.

4. what are nvidia’s competitors doing?

according to the latest report from market research firm idc, nvidia currently controls about 90% of the data center gpu market. faced with nvidia's strong position, many technology giants have not chosen to sit idly by.

dominant cloud computing providers such as amazon aws, google cloud (owned by alphabet) and microsoft azure have increased their investment in developing their own chips in order to gain a favorable position in the future technology competition. at the same time, nvidia's old rivals amd and intel are also not to be outdone and are vigorously promoting their own chip research and development projects.

but for now, these efforts have not yet posed a substantial threat to nvidia’s dominance. amd, for example, expects sales related to ai accelerators to soar to $4.5 billion this year, a qualitative leap from almost zero in 2023. however, it still pales in comparison to nvidia’s forecast that data center sales will exceed $100 billion this year.

5. how does nvidia stay ahead of its competitors?

nvidia has demonstrated an amazing speed in technological innovation and product iteration. it not only continues to upgrade gpu hardware, but also launches highly optimized software support. in addition, nvidia has also carefully designed a variety of cluster system solutions to help customers purchase and quickly deploy its flagship product h100 in batches in a more efficient way, thereby accelerating the implementation of cutting-edge technologies such as artificial intelligence.

in contrast, although intel's xeon processors and other chips also perform well in data processing and can handle more complex data analysis tasks, they are limited by the small number of cores and have relatively slow processing speeds when faced with the massive data processing required for artificial intelligence training. the once dominant data center component supplier has been increasing its r&d investment in accelerator technology, striving to provide customers with more diversified choices besides nvidia products.

6. what is the demand for ai chips?

nvidia ceo jensen huang analyzed the current supply and demand tension in the chip market at the goldman sachs technology summit in san francisco. he said frankly: "customers are deeply frustrated about the shortage of chips. this is because in the current technology competition, every company is eager to be a leader and wants to have the most chips. this undoubtedly makes our customers feel unprecedented pressure and urgency. faced with such a situation, we feel a great responsibility and are doing our best to do our best on the supply side."

huang renxun also said that despite the challenges, the market demand for nvidia's current products remains strong. as supply chain bottlenecks are gradually eased, the newly launched blackwell series chips have received a wave of orders. when asked whether large-scale artificial intelligence spending has provided customers with a return on investment, he emphasized that in today's data-driven era, companies can no longer avoid the trend of "accelerated computing."

7. why is nvidia under investigation?

nvidia's continued expansion in the chip and artificial intelligence fields and its increasingly consolidated dominance have quietly become the focus of attention and potential concern of industry regulators. it is reported that the us department of justice has issued subpoenas to nvidia and several related companies. this move marks that the investigation into nvidia's possible antitrust behavior has entered a new stage and the investigation has been significantly intensified.

although nvidia responded quickly and denied receiving a subpoena directly, it is generally believed in the industry that the u.s. department of justice often collects relevant information and evidence in the form of civil investigation requests, namely the so-called "subpoenas." according to a person familiar with the matter, the core content of this subpoena focuses on the transaction details of nvidia's acquisition of runai and its chip business, aiming to deeply explore whether the transaction involves unfair competition or market monopoly.

facing external doubts and regulatory scrutiny, nvidia said its leading position in the ai ​​accelerator market is based on its superior product performance and technological innovation, rather than any form of coercion or exclusivity. nvidia said that customers have full freedom of choice in the market and can choose the most suitable solution based on their needs and preferences.

8. how do amd and intel compete with nvidia in the ai ​​chip market?

amd, the second largest company in the global computer graphics chip market after nvidia, launched its "instinct" series of products last year to challenge nvidia's market dominance in high-performance computing and artificial intelligence. at the same time, amd and intel are also accelerating the layout of chip designs optimized for artificial intelligence workloads, and claim that their latest research and development results have shown the potential to surpass nvidia's h100 and even h200 in certain scenarios.

however, facing the breakthrough leap promised by nvidia's upcoming blackwell series, its competitors are currently unable to give a comprehensive and specific interpretation or refutation. nvidia's competitive advantage is not limited to its hardware performance, but also lies in its deep technology ecosystem construction. the company pioneered the cuda (compute unified device architecture) architecture, a programming language and platform designed specifically for gpus that can efficiently program and optimize workloads that support artificial intelligence applications. the widespread popularity and application of cuda has invisibly deepened the industry's dependence on nvidia's hardware ecosystem.

9. what are nvidia’s recent release plans?

the most anticipated product is nvidia's blackwell series of chips, which the company said are expected to contribute "substantial" revenue growth this year. however, it is worth noting that nvidia has encountered engineering challenges in the process of advancing the development of this series of products, and this challenge may delay the release schedule of this product.

at the same time, the market demand for nvidia's h-series chips continues to maintain a strong growth trend. as an active promoter of the technology, huang renxun has repeatedly emphasized the importance of accelerating the adoption of artificial intelligence technology, calling on governments and private enterprises to make early arrangements to avoid falling behind in the increasingly fierce competition. nvidia realizes that once customers launch generative artificial intelligence projects based on its technology platform, nvidia will have a significant competitive advantage in future upgrades. (compiled by jinlu)