Cs 1

NVIDIA’s Fastrack
In April 1993, Huang Renxun, who came out of LSI Logic, an integrated circuit
manufacturer, and Chris Malachowsky and Curtis Priem, two young engineers from
Sun, co-founded NVIDIA. Their original intention was to develop a special chip to
speed up the rendering of 3D images in electronic games and bring more realistic
display effects.
After two years of polishing, NV1, the first multimedia accelerator for game
consoles, was launched. This accelerator integrates graphics processing, sound card
and game lever functions. Once released, it attracted the attention of SEGA, the
biggest game manufacturer at that time. However, in the same year, Microsoft
launched Windows 95, which is mainly characterized by graphical user interface
(GUI). The mainstream market of graphics chips has shifted from game consoles to
PCs. Meanwhile, Microsoft has also developed a graphics programming interface,
Direct3D standard, which uses the technology of multilateral image formation. Since
there is no polygon 3D standard like Direct3D and OpenGL (3D drawing
programming interface developed by SGI) in NV1 research and development,
NVIDIA uses quadratic equation texture mapping as its stereo graphics
implementation method, which is incompatible with the industry's general standards,
so its market sales are declining rapidly.
In 1996, NVIDIA stopped selling NV1 and stopped developing NV2, and began
to shift its focus to graphics processor RIVA128. Over the next two years, NVIDIA
launched RIVA128, RIVA128ZX, RIVATNT and other graphics processors. These
new products not only support Microsoft Direct3D and OpenGL standards, but also
exceed competitors ‘Voodoo and RAGE Pro of ATI in terms of energy efficiency.
With their low prices, they have gradually won the favor of the whole plant.
In 1999, NVIDIA 's annual revenue exceeded $150 million and was listed on
Nasdaq. In May of the same year, it sold more than 10 million graphics processors. In
August, NVIDIA launched the first GeForce-named display core, GeForce 256, and
first introduced the concept of GPU. At the end of the next year, with $70 million in
cash and 1 million shares in the company, NVIDIA officially became the industry's
leader by taking the 3Dfx into its pocket.
From 2004 to 2007, the business of NVIDIA games and professional graphics
processors grew steadily. During these four years, David Kirk, the chief scientist of
NVIDIA, was thinking about a longer-term problem: making GPU technology for 3D
rendering universal.
The reason is that Intel's CPU can be shared by all computer applications
through multithreading technology, but GPU can only interact with users through
interfaces such as OpenGL/DirectX. If a suitable programming model can be provided
in GPU and abundant parallel computing resources of GPU can be shared with
developers, then every PC can become a super large-scale high-performance
computer.
The cost of this madness is enormous. Because the relevant CUDA logic circuit
must be added in the design of hardware products, which makes the chip area
increase, heat dissipation increase, cost increase and failure rate increase. At the same
time, it is also necessary to ensure that the software drivers of each product support
CUDA, which is a huge workload for NVIDIA engineers.
From 2009 to 2012, NVIDIA also ushered in the most important period in the
history of development with the power of general GPU based on CUDA in the field of
high-performance computing. In 2012, with the support of general GPU, deep neural
network technology achieved a major breakthrough, including computer vision,
speech recognition, natural language processing, and artificial intelligence technology
in various fields have been rapidly developed. NVIDIA's data center business has also
begun to develop, providing powerful computing support for many companies. It is
certain that in the AI era, NVIDIA has provided strong support for the development
and application of AI technology for all walks of life.
Esports is one of the key drivers of the growth of game business in NVIDIA
emerging markets. Today, some professional game teams even have Championships
based on GeForce GPU standards; a large number of viewers watch E-sports through
online websites such as Amazon and Twitch, and also lay a huge market network for
NVIDIA GPU.
Secondly, the performance improvement of game consoles also creates
opportunities for NVIDIA. Sony PS4 and Microsoft Xbox One have increased the
performance standards of game consoles by 7-8 times. To achieve this level of
performance, users need at least GTX 960 GPU as the standard. In addition, some of
the latest 3A video games have very high-performance requirements. 80% of
NVIDIA's 100 million GPU users need to upgrade their GPUs to play the latest 3A
games, which increases the power and demand of serious gamers to update their PC
performance.
Thirdly, the opening of game content has also driven the demand of NVIDIA high-
performance GPU. In PC games, some developers will give gamers the right to
modify the game and create their own content, which requires high GPU performance,
but also gives NVIDIA the opportunity to sell high-performance GPUs.
Finally, VR will become a new computing platform in the game industry.
Nowadays, more than 90% of PCs are not suitable for VR. It is necessary to upgrade
their GPUs to get a satisfactory VR experience. In addition, the lack of VR game
content also provides conditions for NVIDIA to build a new game ecosystem.

Cs 1

Uploaded by

Copyright:

Available Formats

Cs 1

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Cs 1

Uploaded by

Copyright:

Available Formats

NVIDIA’s Fastrack

launched Windows 95, which is mainly characterized by graphical user interface

there is no polygon 3D standard like Direct3D and OpenGL (3D drawing

programming interface developed by SGI) in NV1 research and development,

NVIDIA uses quadratic equation texture mapping as its stereo graphics

implementation method, which is incompatible with the industry's general standards,

so its market sales are declining rapidly.

launched RIVA128, RIVA128ZX, RIVATNT and other graphics processors. These

leader by taking the 3Dfx into its pocket.

interfaces such as OpenGL/DirectX. If a suitable programming model can be provided

developers, then every PC can become a super large-scale high-performance

CUDA, which is a huge workload for NVIDIA engineers.

network technology achieved a major breakthrough, including computer vision,

speech recognition, natural language processing, and artificial intelligence technology

begun to develop, providing powerful computing support for many companies. It is

and application of AI technology for all walks of life.

Secondly, the performance improvement of game consoles also creates

performance standards of game consoles by 7-8 times. To achieve this level of

the latest 3A video games have very high-performance requirements. 80% of

but also gives NVIDIA the opportunity to sell high-performance GPUs.

Finally, VR will become a new computing platform in the game industry.

their GPUs to get a satisfactory VR experience. In addition, the lack of VR game

You might also like