Computing deals in Openai are the highest dollar

Openai has signed about one dollar in deals this year to calculate the energy to manage artificial intelligence models, and the obligations that dwarf their revenues and raise questions about how to finance them.

Monday Deal with the AMD chips maker It follows similar agreements with NVIDIA, Oracle and Coreweave, where Openai races to find computing power that you think will need to operate services such as ChatGPT.

You will give deals Openai Reaching more than 20 GB of computing capacity, which is almost equivalent to energy from 20 nuclear reactors, over the next decade.

Every 1GW from Amnesty International The computing capacity costs about $ 50 billion to publish in today’s prices, according to the estimates of Openai executives, making the total cost about $ 1 million.

Dealms have linked some of the world’s largest technology groups to Openai’s ability to become a profitable company that can increase its financial obligations increasingly.

“He is not in a position that allows him to provide any of these obligations.”

“Part of the fake Silicon Valley to make it” is to make people enjoy the skin in the game. Now a lot of major companies have a lot of skin in the game in Openai.

Openai burns money on infrastructure, chips and talents, with no place near the capital required to finance these major plans.

The deals also include circular arrangements between the world’s most emerging companies and partners, as well as the complex financing conditions that have not been agreed in most cases.

Openai’s deals with NVIDIA and AMD may cost up to $ 500 billion and $ 300 billion, respectively, according to the Financial Times accounts, although both include incentives that can also help pay the chips that you buy.

The Oracle OpenAi deal will cost another $ 300 billion, while the Data Center Coreave Group has unveiled computing deals with Openai at a value of more than $ 22 billion.

Openai also launched an initiative with SoftBank, Oracle and others in January known as Stargate, which pledged to invest up to $ 500 billion in the US Openai’s US infrastructure. It is not clear how NVIDIA and AMD deal will suit Stargate plans.

The Chatgpt maker did not reveal whether he would buy direct chips or through cloud computing partners, and it is expected to rent some NVIDIA chips.

Openai has obtained great financial incentives from its suppliers for chips purchases.

Nafidia plans to Invest 100 billion dollars in Openai Over the course of the next decade, the provision of Openai’s money can be used to buy NVIDIA chips for its AI’s data centers.

AMD will give Openai’s orders to prevent them to buy up to 10 percent of the company for only a cent per share, depending on its project that reaches certain goals, including some associated with the AMD arrow.

The AMD shares reached nearly $ 204 when the markets were closed on Monday. If they continue to rise, Openai can sell their shares to finance their spending on AMD chips.

“It is a very innovative structure, it didn’t come lightly,” said Lisa Soo, the CEO of AMD on Monday.

The deals with Openai also granted an immediate financial payment to young partners. The market value of Oracle jumped 244 billion dollars after its deal was published last month. AmD shares jumped nearly 24 percent on Monday, which strengthened its market value of $ 63 billion.

Their circular relations added to concerns about a bubble in artificial intelligence, at a time when investors are concerned that spending on artificial intelligence data centers Supporting economic growth for us.

To finance its expansion, Openai raised huge sums of stocks and began to take advantage of the debt markets. She got $ 4 billion in bank debt last year and raised about $ 47 billion of investment capital deals in the past 12 months – although there is a large part of that Difficult negotiations With microsoft, its largest supporter.

People close to the company said that Openai – which is valued at $ 500 billion this month – is also preparing to raise tens of billions of dollars to finance infrastructure.

The risk of credit for the start of the start of anxiety has pushed anxiety. MOODY was placed on the amount of the work center of the future data in Oracle depends on Openai and its non -profitable road.

Openai’s arrangement with NVIDIA for investors is likely to help get more large loans.

Chip Giant, which has recently exceeded 4 meters in the market, has used its routinely mammotitan budget to invest in companies within the supply chain or among its senior customers.

The beneficiaries of these deals use the new liquidity role to buy more NVIDIA chips or borrow money.

NVIDIA has invested in Coreweave, a client and supplier. Coreave also increased more than $ 12 billion of guaranteed debt against NVIDIA chips.

Besides NVIDIA’s support, Openai partners and investors bought their promises to grow in the future and the path of profitability.

Openai expects to double its revenues from its current level of $ 12 billion in the coming years by offering new products and doubling the number of subscribers who pay its primary product.

CEO Sam German said on Monday that he was not profitable, “I was not in my top ten fears.”

He added, “It is clear that we must be very profitable, and we are confident and morning that we will arrive there … Now we are at a stage of investment and growth and if we can provide all this value.”

Openai and its increased number of partners are betting on the use of artificial intelligence will continue to grow significantly. If the growth plateau, or even slow down, the investor’s enthusiasm who has strengthened stock prices against the background of these deals can stumble quickly.

“The company is in a more intense business business than Google or Microsoft at all, it was born without cost discipline,” said one of the ancient warriors in the Silicon Valley.

The investor added that the founder of Amazon Jeff Bezos and the founder of Oracle Larry Ellison “was only found” and the costs of work dramatically reduced, “almost bankruptcy.”

Additional reports. Data is visualized by Eva Xiao and Hahisiang Baladi