Nvidia vs. OpenAI: AI chip market leadership could change

NVIDIA dominates about 80% of the current market for chips designed for artificial intelligence. The sector will experience significant growth, generating revenues of more than $11 billion in 2021.

According to analysis firm AMR (Allied Market Research) this is expected By 2031, sales of these specialized AI chips will reach a value of more than $263 billion,

The company, led by Jensen Huang, stands for ChatGPT, the maker of its chips for artificial intelligence applications, counts OpenAI among its main customers.,

As of March 2023, it was estimated that OpenAI would need more than 30,000 AI chips, likely purchased from NVIDIA, to keep ChatGPT running in the medium term.

Huang’s company’s revenue is projected to increase by 56.4% during 2023, largely thanks to sales of these AI chips.

Despite this excellent moment, the situation may change OpenAI will have plans to develop its own AI chipsIs looking for an alternative to the GPUs which he currently purchases in large quantities from NVIDIA.

According to information from the Financial Times, OpenAI CEO Sam Altman will be looking for investors. Its purpose will be to obtain funding for the development of its own artificial intelligence chips at OpenAI with the intention of reducing the company’s current reliance on hardware supplied by NVIDIA.,

However, to complete this project, OpenAI needs more than funding, You need to cooperate with an IC manufacturer that has the ability to produce semiconductors in the quantities required for your future needs.

The collaboration between Sam Altman’s company and TSMC, although not yet official, appears to be on track.

TSMC, A Taiwanese giant in integrated circuit manufacturingKnown not only for its huge production capacity, but also for its ability to develop and apply the latest technologies in the manufacturing of smaller and more efficient electronic components, known as lithographic nodes.

recently, TSMC and Apple have concluded negotiations that ensure Apple will be one of the first to use TSMC’s new 2-nanometer circuit design when it becomes available.,

The company plans to develop even smaller designs of 1.8 and 1 nanometer. This ability to innovate shows why OpenAI is interested in working with a well-known semiconductor manufacturer.

Although Intel and Samsung are also considered strong competitors, TSMC’s dominant position in the semiconductor manufacturing market with more than 50% market share reflects its leadership.

Although Microsoft already has Copilot, it seems that it will create an AI that will use less energy according to what was recently said in this area.

A major challenge of artificial intelligence today is its high energy consumption.resulting in significant environmental and economic impacts in its maintenance and operation.

because of this reason, Microsoft will look for ways to reduce costs in this technology area Which has experienced rapid growth recently.

Microsoft will work on the development of a more efficient artificial intelligence that will have similar capabilities to the existing one but will require less energy.As revealed by the information.

High energy consumption of advanced technologies is nothing new. This issue was particularly highlighted with the rise of cryptocurrency mining, which used more energy than entire countries.

Artificial Intelligence also faces a similar challenge, as it requires large amounts of energy and water to cool its servers.,

recently, Microsoft has shown its interest in nuclear energy As a solution to meet the energy needs of artificial intelligence, its efficiency was highlighted.

However, This option generates controversy because of the radioactive waste it produces., Nuclear energy is a subject of debate in many countries, with some countries such as France and the United Kingdom adopting it, while others such as Germany and Spain rejecting it.