Video games, the hidden engine of technological development
first level title
01 Jobs and Pac-Man
Friends who are engaged in the electronics industry should know that the integrated circuit industry took off in the 1970s. At that time, Motorola launched the 6502 processor, which was very cost-effective. Although it was only an 8-bit processor, it could already perform complex calculations. Yes, and the price is very close to the people, that is to say, if you make an electronic product and install this 6502 processor, you can sell it to the working class.
How to transform a new technology into a product that the general public can buy, and everyone is willing to buy it, is an important criterion for judging whether an emerging industry can sweep the world, which includes three elements:
If the technology is new;
The technology can already be implemented into a product;
The product is cheap enough;
Lets look back at the development of epoch-making technology products in the past few decades. Do they all meet these three elements: home computers, the Internet, mobile phones, mobile Internet, etc.? Why did we write this series of long articles? I just feel that both AI and blockchain may develop into the epoch-making products mentioned above...
In the year when Motorolas cost-effective 8-bit processor was born, a company and a person were very keen to discover the huge business opportunities that might be hidden in it, so they each chose a direction to spend two branches: the human being looking for I found a good friend and assembled the following ugly electronic device:
By the way, this is the prototype of the Apple Computer Generation (Apple I). The chosen person is Steve Jobs, one of the founders of Apple, and his good friend is Wozniak. The worlds first home personal computer was born in this way. At that time, it was such an ugly duckling, which could be sold for 500 US dollars (according to the conservative data of the US Bureau of Labor Statistics, the purchasing power of the US dollar around 1975 was equal to 6 times that of today, so the analogy is The current $3,000 a set is astronomical). Even more incredible is that 48 years after the birth of the first Apple computer, the market value of this Apple company has reached 2.7 trillion US dollars. If Apple is regarded as an independent national economy, its GDP ranks fifth in the world. Higher than the UK and just below Germany, Japan, China and the US. Jobs established an emerging technology empire.
Stop, stop, what does this have to do with games? Dont worry, I just said that the first direction is the emergence of personal computers for home use, and the second direction is that a company called Atari added the 6502 processor to a game-playing device to develop After two years, finally launched the Atari 2600 game console in 1977, which was the first game console for home use.
The console brought in $2 billion in annual revenue in 1980 (equal to $12 billion in todays dollars). Todays players are completely indifferent to Ataris brilliance, but the one platform, multiple games model it pioneered is still being applied until now: the host company develops, and the games in the cartridge can be handed over to the game studio. The last level to pass in the movie Ready Player One is the Pac-Man game played on the ice, which exploded on the Atari 2600 game console.
Ataris sudden wealth made many technology companies at that time suddenly see a new direction for the rapid transformation of technological capabilities into wealth, so Sega and Nintendo, which are still famous until now, began to enter the market. In 1983, Nintendos red and white machine (FC game console: Famicom) swept the global game market like a tornado, and also brought the most indelible childhood memory of our generation, Plumber: Brother Mary. The red and white machine was not discontinued until 2003, and a total of more than 60 million units were sold. The Xiaobawang learning machine that we are familiar with in Dongfang is the imitation disk of the red and white machine.
Gossip here again, the first direction mentioned above is also learned by Jobs from Atari. Jobs, who has been a boss all his life, has the only part-time job resume that he worked in the game console team of Atari Corporation. During his part-time job, he learned the powerful functions of the 6502 chip, and then he went out to start a business and created an Apple.
At the end of the 1980s, the competition between Sega and Nintendo was fierce, and the computing power of 8-bit processors was not enough, so we made 16-bit ones! In this way, under the pressure of terminal products and market demand, the research and development of chip technology has advanced by leaps and bounds. Whoever develops it first will make money. After all, this transformation path has been established.
first level title
02 Magical Nvidia
With the rise of this wave of AI boom, the name of NVIDIA (NVIDIA) has been frequently reported by the media. Friends who have experience in building computers, as well as big Vs who have become rich through mining, should be no strangers to NVIDIA. It is The worlds largest graphics card manufacturer. High-end 3A games need a graphics card, and mining also needs a graphics card (in fact, what is needed is the GPU on the graphics card, and the computing power of the GPU exceeds the CPU in some aspects), but what does it have to do with AI? Lets just sort it out.
Video games have been developed for nearly fifty years, and players have higher and higher demands for game images and videos, so the computing power of the GPU on the graphics card is constantly iterating and evolving; later, some extension industries, such as video capture and encryption industries, discovered The characteristics of GPUs high computing power have given new empowerment to graphics cards; around 2000, AI entered the stage of deep learning (for details, please refer to the first article in this series of long articles, Xinton created a way to describe neural networks ——deep learning), scientists have found that the calculation method of the graphics card is especially suitable for the learning of neural networks, that is, the more information and data fed, the better the effect. It is said that OpenAIs breakthrough came from an internal test that used tens of thousands of Nvidia GPU A 100, and then, like Gods golden finger, a new world began to reveal a mysterious corner.
As a result, graphics cards have become a must for the growth of AI companies. If you look at the stock price of Nvidia in the past two decades, it really took off around 2016. In the previous article, you know that OpenAI was established in 2015. Nvidia has completed its counterattack with the tide of AI: the current market value is about 680 billion US dollars, far surpassing other chip companies Intel and AMD, and has become a magical technology bull stock second only to Apple, Microsoft, and Google.
The current mainstream GPUs that provide computing power for large AI models are two models produced by Nvidia: A 100 and H 100 . Some analysts even regard the number of A 100 and H 100 owned by AI companies as one of the important indicators of the enterprise, that is, English content. High English content is like taking Viagra, and it will soon develop... This Its the definition of NVIDIA that I interpreted, hahaha, its messy.
Even Musk recently announced that Twitter has purchased about 10,000 Nvidia GPUs. Just when everyone is envious of Nvidias good luck, we cant ignore the fact that as early as 2016, Nvidia founder Huang Renxun donated an AI supercomputer DGX-1 to OpenAI to demonstrate Nvidias GPU With unique advantages in AI computing power, the computing power of this special machine can already compress OpenAIs one-year training time to one month. Therefore, it can be seen that Nvidia has long been laying out the AI track. A company with a market value of 680 billion US dollars will not come from the sky. The strategic decision-making and business negotiations behind the scenes should be very exciting. interpretable.
So how did Nvidia come about? Why can it span several industries including video games, industrial video, encryption mining and AI computing power? Lets continue the gossip:
In fact, like the Internet, the integrated chip was originally born out of military needs. In the 1950s, the U.S. Department of Defense hoped to add a control unit to missiles and other weapons to hit targets more accurately, so they found a few universities. The laboratory offered a reward for research, and Shockley, a famous scientist at MIT (Shockley is known as the father of the transistor and won the Nobel Prize in Physics), led a group of young talents to come up with the concept of the chip.
Shockley has long been successful and famous. After making some achievements, his biggest hobby is to go to AMA every day, enjoying flowers and applause, and saying to the outside world that the project party is doing something. In fact, he has been spending money from the investor, the US Department of Defense. But the young talents under him really want to do things to realize their dream of getting rich. They have been demanding to increase research and development, reduce costs, and scale production, so that they can break through into large civilian fields and dig and dig and earn a lot. money, but Shockley just refused to let him do it.
Later, eight young people emerged from the Shockley Laboratory, and a new company called Fairchild was established, known as the Eight Fairchild Rebels. The most famous of them is Moore, who defined Moores Law.
The covenant of eight people is also very dreamy. They took a piece of US dollars and signed everyones names. The meaning couldnt be more obvious: Lao Xiao, dont make big cakes for us anymore! What I want is green banknotes!
In fact, in the history of technological development, this kind of gang warfare can often bring subversive revolutions, from the early Fairy Boys and Eight Rebels who led the chip revolution, to the Paypal gang in early 2000 (members founded Tesla, LinkedIn, Youtube and other companies) launched the Internet revolution, and then to the founding group of OpenAI mentioned in the previous article in 2015.
Speaking of Fairchild, this company is epoch-making:
First, it defines the semiconductor industry in terms of technology, and the entire research and development idea has turned to silicon as a material, which can mass-produce integrated circuits;
Second, it defines the business model of venture capital + incubation company. Fairchild Eight Rebels first approached Arthur Rock, the father of the venture capital industry on Wall Street, as an FA, and introduced the funder’s father, Sherman Fairchild, who is the founder of Fairchild. The name comes from the name of the gold masters father. This kind of kneeling and licking is effective. This Sherman Fairchild has a lot of background and is the largest individual shareholder of IBM. Fairchilds first order after its establishment came from IBM. Since then, this type of venture capital business model has become popular and has been popular until today, that is, investors do not just give money, but get involved in company management and operations, and use their own resources to help entrepreneurs succeed. In short, more in-depth (This is what W Labs is doing, hahaha).
Third, it promoted the emergence of Silicon Valley. Fairchild gradually became the Whampoa Military Academy of the semiconductor industry. After several years of cooperation with Fairchild, the eight rebel generals of Fairchild, they did not agree on the terms later, and they also came out to do it alone, and led a group of other colleagues to come out. . Everyone started a new company, scattered around the Fairchild office, and this area became what is now Silicon Valley. Among these new companies were Intel and AMD. The young people who left Fairchild have founded more than 100 semiconductor companies, and Fairchild has really become a fairy incubator.
Speaking of returning to Nvidia, people from Fairchild founded AMD, and Nvidia’s boss Huang Renxun was an engineer for AMD, and later came out to start a business and started Nvidia. These companies and people are inextricably linked. It is said that Huang Renxun and Su Zifeng, the current speaker of AMD, have an indirect relationship.
The idea of Nvidias establishment is: you Intel and AMD go to PK CPU, I specialize in GPU. Anyway, computer games are getting better and better, and the computing power of traditional CPUs is getting more and more difficult. I specifically separated the graphics processing part to make a new chip, defined it as a GPU, and installed it on the graphics card. This was still a blue ocean at the time.
In the early days, the game industry contributed 100% of Nvidia’s revenue. After making money, Huang Renxun crazily increased his research and development efforts and expanded GPUs to industrial and other fields. The law is very similar to Huawei: research and development first, followed by the market. The graphics processing segment is becoming more and more popular. In addition to AI, even drones on the battlefield use GPUs for graphics recognition processing.
So taking Nvidia as an example, games, as the most resource-hungry software, reversely force the improvement of hardware performance. At the same time, games are the best industry to link B-side and C-side. Users are willing to go to Pay the bill.
So far, I have successfully asked the question again. Games are the driving force behind the progress of science and technology in the past half century.
To be continued. This series is collectively created by the W Labs AI Chain Games Research Group. Thanks to the team members Gua Ge, Jia Ran, Bao Bao, Brian, Xiao Fei, and Hua Ge for their hard work! The main writer of this article is Gua Ge.
Welcome to reprint, originality is not easy, please indicate Guadain Laboratory W Labs produced
Guadain Laboratory Website: http://www.wlabsdao.com/
