Whatever the fate of this company in the future, right now it is writing not just its history, but also, it seems, helping humanity move to a new stage of development.
But just recently, and for the majority, let’s not lie, and today Nvidia remains simply largest graphics chip developer. This is not Google, not Apple, and not the new-age OpenAI. But the world is changing dramatically, and this cannot happen without the active participation of Nvidia itself.
This article will help you understand how the company managed to change its place in the second tier to the role of the only engine of the technology industry.
Short description
Nvidia is a $3 trillion company.
Why Nvidia started the development of II
How Nvidia helped accelerate the development of neural networks
The evolution of specialized servers from Nvidia
Where are the competitors?
Nvidia and OpenAI: friends or not anymore
What’s next
Nvidia is a $3 trillion company.
Nvidia office in California.
Nvidia made a name for itself back in the 90s by creating graphics chips. Who among the gamers has not heard about the “red versus green” war, which has been going on for two decades with varying success. In this market, Nvidia essentially has one competitor – AMD, which bought out ATI.
The mining boom caused a frantic demand for video cards, which brought super-profits to the manufacturer. Nvidia’s stock price increased 6-fold from 2019 to 2022, while AMD’s stock rose “only” 5-fold. It is not surprising that with a sharp drop in mining profitability, there is an oversupply of video cards on the market. And analysts sharply adjusted the views of graphics chip manufacturers.
But if AMD shares, after the expected fall, resumed their smooth growth, then something completely unexpected happened to Nvidia.
In terms of capitalization, Nvidia has overtaken Apple!
In 2024, green stocks rose sharply in price. Nvidia itself reached a capitalization of 3 trillion dollarsbecoming second in the world in terms of indicator and second only to Microsoft.
What happened? The reason is clearly not gamers. It’s all about Nvidia’s close connection with modern AI and machine learning technologies.
◦ Back to contents ◬
Why Nvidia started the development of II
The legendary GeForce 8800 based on the G80 chip.
Back in 2006, Nvidia introduced the next series of its graphics accelerators based on the G80 chip. It was the first to support CUDA technology. And this, as it turned out later, played a vital role in the development of AI. What’s the matter?
Video cards of that generation received ten stream processors, an infrastructure of up to 128. These cores switch to work with a multi-threaded algorithm. They themselves weren’t as versatile as a processor core, but there were many, many of them. How CUDA kernels limit the ability to display graphs is another story, but they quickly found a new field of application.
The unified architecture makes it noticeably more difficult to solve any parallel problems. As a result, the video card may turn out to be complex and no longer associated with graphical computing in various fields of science. The processor of our home computer usually has 4-8 cores, but, for comparison, on a fairly budget RTX 4060 video card there are more than 3000 of them!
◦ Back to contents ◬
How Nvidia helped accelerate the development of neural networks
OpenAI emerged in late 2015 with the goal of openly exploring AI. And already on April 27, 2016, a public beta version of the OpenAI Gym platform was presented, which developed and compared machine learning algorithms.
And just 4 months later, the company received an Nvidia DGX-1 supercomputer as a gift. Jensen Huang personally presented the “toy” to Elon Musk at the OpenAI office.
A gift from Nvidia was for the OpenAI event
The computer had 8 graphics cards with CUDA support and cost 129 thousand dollars. DGX-1 was designed to solve the problems of training artificial intelligence models. The processing time for a typical data set has been reduced from 6 days to 2 hours!
OpenAI received official tools for its developments, and Nvidia received a new profitable use for its developments. At that time no one realized what this could lead to.
◦ Back to contents ◬
The evolution of specialized servers from Nvidia
The company didn’t stop at creating DGX-1 – appetite comes with eating! Each new generation of dedicated servers literally doubles system performance to improve the computing design, increasing the number of GPUs, increasing their memory and limiting bandwidth.
Comprehensive improvement gives explosive results! And Nvidia was able to provide such high-quality jumps literally once every 1-2 years.
The new architecture brings big gains. And all this is developing rapidly!
Server A100 seems to be tired already
GPU memory growth on Nvidia servers
How in demand are Nvidia servers? Very! The largest companies are participating in the race, trusting their AI for their services and pumping up their neural networks.
Who has Nvidia’s H100 GPU?
◦ Back to contents ◬
Where are the competitors?
It is assumed that there are two major players in the graphics accelerator market. But why is Nvidia so in demand for neural networks, and not AMD? There are several reasons for this:
1. CUDA platform. In fact, it has already become the standard for working with a graphics processor. Developers have received many libraries and tools for developing, optimizing and deploying applications using AI and machine learning services. CUDA has support for popular frameworks for developing neural networks: TensorFlow, PyTorch, Caffe and others.
2. Custom GPU architecture. Nvidia was able to create a whole series of processors: V100, A100, H100, which are specifically designed for high-performance tasks in the field of AI. There are special Tensor Cores that make it difficult for neural networks to work.
The leader is obvious. And the situation has not changed in a year. According to https://wccftech.com
3. Hardware solutions. Nvidia offers comprehensive hardware solutions based on DGX video servers, which are designed specifically for solving deep learning problems and switching computing system performance.
4. Pioneer Bonuses. Thanks to the company’s management’s interest in AI, the company became the first leading manufacturer of specialized hardware in this market. Nvidia quickly built its ecosystem, which quickly filled with developers and researchers. As a result, technology has become even more popular and accessible.
Of course, AMD is not sitting idly by, watching the division of a large “pie” of a rapidly growing market. Thus, at the end of 2023, the company introduced its specialized software accelerator Instinct MI300x, a direct competitor to Nvidia’s H100. Each company has this superiority in normal tests, blaming competitors for bias. At the same time, the market is clearly not ready for drastic changes, trusting the leader.
AMD managers demonstrate the superiority of their products. Believe?
◦ Back to contents ◬
Nvidia and OpenAI: friends or already competitors
Let’s say Nvidia is a leading technology manufacturing company, and OpenAI is developing algorithms in the field of AI, creating language models. Some give iron, and the second – soft. OpenAI, like other developers in this area, creates its product based on Nvidia processors. Both companies, working closely together, help each other grow.
And although Nvidia and OpenAI do not appear to be competitors, as the industry develops they are increasingly beginning to compete for customers. In February, Nvidia launched its own chatbot, Chat with RTX, designed to run on Windows computers with GeForce RTX 30-series or 40-series graphics cards with 8+ GB of video memory. And at the same time, even the Russian language happens!
The special thing about this chatbot is that it can run locally on a PC and provide answers based on the user’s documents and videos. Chat with RTX via the link gives a description of the video on YouTube, processes PDF files, which makes it faster than cloud-based ChatGPT or Copilot. Nvidia gives out a demo version of the product for free – an excellent advertisement for its capabilities. You can chat locally, without the Internet!
Nvidia chatbot transcribes YouTube video
The company’s CEO, Jensen Huang, said ambitiously: “We have the capabilities, scale and reach to help any company in any area of professional artificial intelligence.”
And words are not wasted because of this. Last year alone, Nvidia invested more than $2 billion in 35 AI startups.
Jensen Huang donated a new DGX GH200 server based on OpenAI. There is no such thing as too much PR!
OpenAI is not sitting idly by either. The situation when one of the manufacturers on the market produces more than 80% of the required chips is not the best for the customer. It is no coincidence that Sam Altman is trying to launch a project on 7 trillion dollars to cover the shortage of computing power and AI development.
◦ Back to contents ◬
Nvidia: what’s next?
Nvidia stock’s rally comes from very low levels with a bright future ahead. It’s no secret that AI technology is experiencing a real boom! This is of interest to everyone, the scope of their application is already clear, and how many more things can you come up with! All you need is computing power and all hope lies with Nvidia. The company is flirting with OpenAI, but possible clients are still willing to wait their turn, such giants as Microsoft, Meta and Google. At one time, even Yandex suffered from a shortage.
The server market is growing, but mainly due to interest in Nvidia solutions
Remember the news that Microsoft is going to launch a $100 billion supercomputer for OpenAI? Well, who will provide the server if not Nvidia? Not even futurologists, AI experts predict the emergence of AGI (artificial intelligence at a level not lower than human) in 2027-2028. All that remains is to assemble a single computing cluster for just a trillion dollars. Big tech seems to have little money for this. And Nvidia is an equal partner project here.
Nvidia is an indispensable part of the revolution currently underway. The company has paved the way for future sales of powerful computers. And on this soil, the tree of the future unexpectedly quickly grew, the fruits that we are already beginning to use.
Source: Iphones RU
I am a professional journalist and content creator with extensive experience writing for news websites. I currently work as an author at Gadget Onus, where I specialize in covering hot news topics. My written pieces have been published on some of the biggest media outlets around the world, including The Guardian and BBC News.
