TechCrunch commented that the AI Data Centre Financing Scheme, published this week by Ingweida, ostensibly introduced up to $500 billion, and that what is really more noteworthy is the company ' s attempt to create a more stable secondary market for the old GPU. If established, the arrangement would not only continue to drive a new round of calculus-building, but could also change the pricing of AI hardware.

Young Waida guarantees GPU residual value

According to the article, Apollo, Beled, Blackstone, Brookefield, Goldman Sachs and KKR are willing to commit up to $500 billion for the construction of the AI data centre. In order to attract these long-term capital, Weida agreed to use its own funds to partially guarantee the GPU residual value of related transactions.

Specifically, if the GPU as collateral were to be disposed of at a lower price than was expected on the books, the difference of up to 25 per cent would be covered by YVD. That is to say, if the data centre owner defaults and the lender does not sell the equipment at the expected price, it will have to bear part of the loss.

The core goal is to make big used hardware markets.

According to the article, this design is not just a financing tool, it's more like the longer life cycle that Britain is looking for for old AI chips. For start-up companies, business clients and research institutions, hardware options are more likely to be available if the second-hand GPU market becomes more active and the threshold for gaining credit is likely to fall.

CEO Hwang In-hoon described the AI server as a long-term investmentable infrastructure rather than a fast-depreciated device like a personal computer. His judgement is that the need for computing power will flow between different customers, cloud service providers and operators, which will allow the Yvesta computing platform to have a metronom at all times, thus supporting the residual value of the equipment.

The risk is that responsibility rises when demand weakens.

The article also notes that the most sensitive aspect of this model is what financial markets call “reverse risk”. If AI's demand weakens and GPU prices fall, the compensation burden that Britain has to assume will increase, and this will often be accompanied by corporate revenue constraints.

Market vigilance is due to the fact that this arrangement is easily reminiscent of telecommunication technology during the Internet bubble. The latter had promoted the sale of equipment by providing financing to clients, which had eventually been severely damaged after the industrial bubble burst. In-hoon has also recently explained in social platforms and television interviews that, instead of taking the main financial risk on his own, Wei Da this time introduced independent long-term institutional capital and that companies only partially protect future values.

The article also mentions that Young Weida has previously invested billions of dollars in client-related transactions such as OpenAI, Anthropic and CoreWeave, Nebius, Firumus and Lambda. Bloomberg had previously estimated that British Wida was moving forward this summer with similar revolving financing arrangements totalling approximately $750 billion.

If the mechanism works well, Ying Wei Da will find new sources of financing for the AI data centre. This is due to the fact that traditional funding is becoming more laborious, with some of the large technology companies already carrying more debt, increasing stocks or continuing to consume cash. In turn, if AI uses growth to slow down, or new technologies to make existing infrastructure more obsolete, the pressure on the model will increase rapidly.