"Equity for procurement" -- The agreement between AMD and OpenAI is "rare in semiconductor history."

"Equity for procurement" -- The agreement between AMD and OpenAI is "rare in semiconductor history."

```

AMD and OpenAI have announced a GPU supply agreement worth up to $90 billion, using an unprecedented "equity for purchase" model that transforms traditional hardware supply into an equity incentive mechanism. This structural innovation could redefine the financing methods of AI infrastructure.

According to Hard AI, on October 6th, OpenAI announced plans to purchase and deploy up to 6 GW of AMD Instinct series GPUs, with potential sales reaching $90 billion. In exchange, AMD issued OpenAI warrants to purchase up to 160 million shares of AMD stock at an exercise price of $0.01 per share.

According to calculations, if AMD's stock price rises to $600 in the future, OpenAI's potential 160 million shares would be worth $96 billion, roughly matching the value of the hardware involved in the transaction. As one user commented:

It seems that this deal suggests that if AMD stock reaches $600 and GPU deployment milestones are met, OpenAI effectively gets 10% of the shares for free.

AMD CEO Lisa Su called it a "win-win," while OpenAI CEO Sam Altman emphasized the necessity of "massive computing power" to realize AI's potential.

Analysts believe that on the surface, this is just a supply agreement. But in essence, it can be seen as a financial instrument, turning AI hardware sales into equity allocation and directly linking AMD's long-term valuation to OpenAI's infrastructure growth.

Even accounting for equity dilution, the enterprise value AMD gains from the 6 GW demand far exceeds the impact of the additional share issuance. After the announcement, AMD's stock price soared nearly 24% on Monday.

(AMD stock once opened above $220)

Procurement in Name, Deep Binding in Essence

Analysts believe the brilliance of this agreement lies in its financial structure.

AMD's 8-K filing shows that the warrant is structured to function as a performance-based equity incentive rather than traditional equity dilution, and AMD does not have to give up governance or board representation rights.

The nominal exercise price of $0.01 per share means that once the exercise conditions are met, OpenAI can almost obtain AMD stock at zero cost, with its return hinging entirely on two core pillars: operational execution and market performance. Specifically:

The first portion of equity is tied to the deployment of the initial 1 GW MI450 series GPU batch, which begins in the second half of 2026; once the hardware is delivered and accepted, this portion of equity vests.The rest of the equity unlock is tied to two dynamic variables—OpenAI's subsequent GPU purchase volume and AMD's stock price.

This design creates a powerful symbiotic relationship. OpenAI can only obtain full equity benefits if it actually purchases and deploys AMD hardware, and this move effectively boosts AMD's data center business growth, further recognized by the capital markets (reflected as a rising stock price).

The agreement is valid until October 2030, and unvested shares will automatically be canceled. AMD also retains anti-dilution and transfer restriction clauses to ensure OpenAI cannot casually sell or transfer the warrants.

The Win-Win Calculation Between AMD and OpenAI

From a financial and strategic perspective, this agreement creates significant value for both sides.

For AMD, this is an innovative way to acquire customers. It turns the traditional upfront discount or rebate costs into equity costs tied to future performance, locking in at least 1 GW of guaranteed revenue and tying the risk of equity dilution to actual business growth.

This creates a key flagship customer and market validation opportunity for AMD’s next-generation data center GPU—the MI450 series and its Helios rack-level platform.

For OpenAI, the agreement solves two major strategic challenges at once.

First, it secures a stable source of non-NVIDIA hardware in a tight supply market, achieving supplier diversification.

Second, it creates a potential "self-financing" path. As purchase milestones are achieved and AMD's stock price rises, the value of OpenAI's vested shares increases accordingly; the proceeds can be used to fund future GPU purchases, effectively offsetting the huge capital expenditure pressure of AI infrastructure.

A New Industry Paradigm: Reshaping AI Infrastructure Financing

The deal’s far-reaching significance is that the computational power for AI is evolving from a pure capital expenditure into a financialized, securitized asset class.

Analysts believe it also clearly illustrates a different ecosystem construction paradigm compared to industry leader NVIDIA:

NVIDIA's Stargate and other projects rely on direct joint investments with giants like Microsoft and Oracle, locking partners into its vertically integrated infrastructure through large-scale capital deployment.Intel splits the financial pressure of fab expansion by bringing in joint investments from NVIDIA, SoftBank, and the US government, blurring the lines between suppliers, customers, and even competitors.

AMD's model is different. It does not require partners to directly co-invest, but retains the core of the commercial transaction, i.e., OpenAI still pays in full for hardware.

But by adding equity incentives, AMD cleverly transforms the customer's "consumption behavior" into an "investment behavior," encouraging customers to expand their purchases and share in the supplier's growth dividends.

Analysts point out this model may be more attractive for AI companies and cloud service providers without the capital strength of top giants but still looking to form deep alliances with semiconductor partners.

Execution Risks and Market Prospects

Despite the rosy outlook, this grand agreement still faces many risks and uncertainties.

First, some details of the agreement are not transparent. AMD's public filings omit key equity vesting schedules and specific technical trigger conditions, making it difficult for outsiders to predict revenue recognition and equity vesting pace precisely.

Second, the 6 GW target is a cap, not a hard commitment. Currently, OpenAI's binding commitment covers only the initial 1 GW deployment, with subsequent purchases entirely dependent on future business needs and performance. If OpenAI’s expansion slows down, the subsequent equity incentives may never be realized.

Finally, execution risk is AMD’s biggest challenge. Delivering GPU clusters of this scale requires AMD to maintain a stable supply chain over several years, including wafer foundry capacity, substrates, and high-bandwidth memory (HBM) supplies.

According to analyst conference calls, AMD has invested in supply chain arrangements for years, but any disruption could cause delivery delays. In addition, physical deployment is also limited by transformer delivery cycles, power grid approvals, and other factors, which are common bottlenecks for the entire industry.

This article is from WeChat Official Account "Hard AI". For more AI insights, click here

Risk Warning and DisclaimerThe market carries risks, and investments should be made cautiously. This article does not constitute personal investment advice, nor does it take into account individual users’ specific investment objectives, financial situations, or needs. Users should consider whether any opinions, viewpoints, or conclusions in this article suit their own circumstances. Invest at your own risk. ```