文 | 半导体产业纵横
10月28日，谷歌母公司Alphabet公布三季度财报：经营活动带来约423亿美元现金，购置设备和物业却支出约514亿美元，自由现金流为负约91亿美元。搜索、广告和云业务仍在增长，但当季经营产生的现金，已经不足以覆盖资本投入。
对CoreWeave、Nebius这类专门采购GPU、向外出租算力的新云服务商，资金更直接地决定扩张速度。10月22日，黄仁勋谈及这一群体时表示，算力需求极其旺盛，服务能力也很强，缺少的是营运资金；AI多个领域同时扩大投入，资金需求已超过可获得的资本。
三天后，马斯克给出了更悲观的判断。他表示，过去几周，债务市场几乎不再向缺乏AI数据中心以外大规模收入的公司提供资金，并预言两百多家新云将经历大规模倒闭与整合，只有少数能够生存下来。黄仁勋相信资金问题能够得到解决，马斯克却认为，不少企业可能等不到那一天。
一边是持续增长的算力需求，另一边是越来越受关注的投入与偿债压力，芯片的大买主们，真的快没钱了吗？
订单涨得快，现金却没有跟上
过去几年，CoreWeave、Lambda等新云通过GPU抵押贷款扩大集群，资金提供方也从私募信贷机构扩展到大型银行。运营商以设备和客户合同作为融资依据，提前取得采购资金，再用算力租金偿债。这种安排支持了快速扩张，也使下一轮采购越来越取决于贷款人如何评估设备价值和客户信用。
10月初，英伟达推动的算力融资项目中，部分银行和资管机构要求比原先方案更强的保障。贷款人关心的不只是GPU今天能否租出去，还包括几年后能赚多少钱、客户付款能否持续，以及违约后设备还能卖多少钱。
双方对GPU价值的判断已有分歧。Impax资产管理公司的高级投资组合经理Tony Trzcinka指出，银行通常按3至4年的折旧安排评估GPU，而英伟达认为高端GPU可以创造长达十年的收入。设备还能使用多久，与它能否持续产生足够收入，并不是同一件事。对缺少其他业务收入的新云，抵押物估值更保守、担保要求更高，都可能意味着需要拿出更多自有资金，或接受更贵的贷款。
CoreWeave在债券市场也遇到了压力。公司6月发行的一批债券，发行时收益率约为9.7％，到10月中旬，随着交易价格下跌，市场收益率已升至约15％。投资者要求更高回报来承担其信用风险，公司后续发债或再融资，也可能需要接受更高的融资成本。
其二季度账目显示，经营活动带来的现金净流入约6.79亿美元，购置设备和物业的现金支出却达到64.22亿美元。同一季度，净利息费用约6．4亿美元，相当于营收的四分之一。公司既要为客户预订的算力购买设备，也要承担已有借款的利息，持续扩张不能只靠当季经营所得。
10月20日，CoreWeave回应债务疑虑时，强调多数债务有长期客户合同支撑。公司先取得信用较强客户的采购承诺，再据此融资、建设基础设施；这些合同的约定收入足以偿还相关本息，不必依靠客户续约。其二季度末待履行合同对应收入达到1040亿美元，三季度初又取得250亿美元新增承诺。
已有合同为相关债务提供了还款来源，但新增采购仍需继续筹资。如果贷款人要求更高回报、更多担保，运营商即使拿到新订单，也要重新计算扩大集群是否划算。新云面临的压力，由此从投入规模大，进一步延伸到每一轮融资的成本和条件。
这些采购计划能否持续，也进入了台积电的产能评估。魏哲家在10月15日法说会上表示，公司会考虑云厂商资本开支、财务状况等因素，并与终端客户交流。晶圆代工厂的直接客户是芯片设计商，但先进制程和封装产能需要提前安排，设计商提出的需求，最终仍要由终端买家的投入支撑。核查这些买家的财务状况，有助于台积电判断需求能够持续多久。
芯片厂商也开始参与筹钱
英伟达已经在一些项目中提供了支持。10月28日披露的交易信息显示，印尼算力平台Zankore正在为31亿美元GPU贷款引入更多参贷银行。它与英伟达签订的六年协议，包含收入分成和信用支持，涉及首批100MW设施中尚未售出的算力容量。相关融资在9月已经公布，近期披露使供应商参与其中的方式更加具体。
按英伟达此前介绍的一般模式，公司会为部分算力容量提供最低收入承诺，让金融机构有依据支持项目建设。新云仍可将算力出售给第三方，满足条件后，英伟达参与收入分成。运营商获得了自身信用之外的支持，英伟达则在硬件销售之外，有机会分享后续租赁收益，同时承担相应的支持责任。
需要筹措设备资金的还有模型公司。在定制芯片一端，博通已经参与到Anthropic的资金安排中。10月26日，博通与金融伙伴宣布一项630亿美元融资交易，由新设的项目公司购买其定制AI加速器及网络设备，再出租给Anthropic，预计2027年开始交付。银行和黑石等机构提供前期采购资金，模型公司通过租用取得设备，博通按计划组织供货。
为了使这项安排成立，博通提供了设备残值支持。10月27日提交的监管文件显示，全部机架部署后，其最高潜在责任约430亿美元，按未折现口径计算。这项责任随机架交付而增加，随Anthropic支付租金而减少；如果客户违约，博通可能承担约定未偿款项与设备处置价值之间的差额。
430亿美元是潜在责任上限，而非博通已经支付的现金。它所对应的变化却很具体：供应商完成设备销售之后，仍需关注客户能否支付后续租金。金融机构提供资金，客户分期承担设备成本，博通则以部分风险承担换取采购推进，芯片供应与融资安排由此联系得更紧。
博通支持的是设备价值，英伟达在上述模式中支持的是部分算力收入。两种做法都让贷款人多了一层保障，也让一些难以独立承担巨额投入的客户有机会扩大采购。对供应商而言，争取这类订单，除了产品性能、交付和价格，还涉及愿意为客户提供多少支持。
但贷款人仍有自己的判断。10月初，部分参与英伟达相关融资项目的银行和资产管理机构，要求比最初方案更强的保障。供应商参与并没有让风险消失，而是改变了风险由谁承担。交易能否继续扩大，要看金融机构是否接受这些条件，以及芯片厂商愿意为新增销售保留多少后续责任。
资金仍在进入，只是条件各不相同
马斯克看衰新云融资的第二天，SpaceX便宣布获得约400亿美元GPU融资租赁承诺。由Apollo管理的实体购买GPU及相关设备，再出租给SpaceX，预计2027年上半年随设备交付分批起租。融资仍需满足约定的交割条件，但已承诺的规模表明，大型GPU采购仍能吸引金融机构参与。
马斯克解释SpaceX的融资优势时，强调了投资级信用，以及星链和发射业务收入。它与主要依靠GPU租金偿债的新云，能够向贷款人展示的资金来源并不相同。看衰部分运营商的融资前景，同时为自身扩大采购筹资，并不矛盾；差别在于谁有条件承担持续投入。
成熟业务提供的现金，也让大云厂商保有更大余地。微软10月28日披露，截至9月底的季度经营现金净流入为580亿美元，购置设备和物业的现金支出为394亿美元，自由现金流仍有186亿美元。公司预计2026自然年资本开支约1700亿美元，并预期2027财年自由现金流保持为正，同时表示会在条件合适时使用额外融资。
此前提到的Alphabet，则在同一季度通过普通股和高级无担保票据净筹资约411亿美元。经营现金流没有覆盖当期资本投入，并没有阻止它继续从资本市场取得资金。对这类买家，筹钱是安排巨额投入的一部分，不能仅凭融资动作判断其采购能力正在衰退。
新云并非都能调用这样的业务收入，但其中一些企业已经凭大客户合同取得支持。10月22日，Nebius与花旗签下35亿美元融资额度，用于数据中心建设、GPU采购等。公司表示，手中超过400亿美元、来自投资级客户的长期合同，可以支持后续资产融资，相关客户包括微软和Meta。
这也呈现了大云厂商与新云之间的另一层关系。微软、Meta既自行投入基础设施，也向外部运营商购买算力。新云替客户承担硬件采购和运营，再用这些客户的多年付款承诺争取融资。贷款人看中的，不仅是新云自身的规模，还包括租约另一端是谁、合同收入是否可靠。
Nebius此次融资期限为364天，将与客户预付款、运营现金流搭配使用，支持近期扩张。2027年及以后的长期资金仍需继续安排。大客户合同没有一次解决所有资金需求，却为分阶段采购提供了条件，也使其筹资能力与缺少稳定租户的运营商有所区别。
10月27日，HIVE旗下BUZZ Alpha Cloud也公布了约1.29亿美元、固定年利率9％的三年贷款，用于采购GB200集群，计划以Cohere三年合同的收入偿还。贷款约覆盖购置成本八成，余下出资来自此前可转债融资所得，设备预计年底前交付。
有客户、有匹配期限的合同，运营商同时承担部分出资，这笔采购因而有了具体资金安排。9％的固定利率也提醒市场，筹得到钱与资金便宜是两回事。新云需要比较的不只是GPU租金和设备价格，还包括借款成本，以及这些成本如何影响下一批采购能够承受的规模。
结语
芯片的大买主们并没有一起陷入资金枯竭，但持续扩大采购，正在考验各自的筹资能力。成熟业务收入、长期客户合同，以及供应商提供的支持，决定了不同买家还能以什么条件继续投入。对芯片产业而言，需求有多大之外，还要看谁能按计划付款、完成采购，以及这种投入能够持续多久。
本文由 AI宝库自动抓取并翻译，版权归原作者所有；如需查看原文请点下方链接。
The semiconductor industry
On 28 October, Alphabet, the Google parent company, released a three-quarter financial newspaper: operating with about $42.3 billion in cash, purchased equipment and property with about $51.4 billion in expenditure and a negative $9.1 billion in free cash. Search, advertising and cloud operations continue to grow, but the cash generated by the current season is no longer sufficient to cover capital inputs.
For new cloud service providers, such as CoreWeave and Nebius, specializing in the procurement of GPUs, and in off-site leasing of computing power, the speed of expansion is more directly determined by the funds. On 22 October, Hoang In-hoon, referring to this group, stated that there was an extremely strong need for credit, a strong capacity for services and a lack of operating funds; and that there were multiple areas of AI that were expanding their inputs and that the need for funds exceeded the available capital.
Three days later, Mask made a more pessimistic judgment. He stated that in the past few weeks the debt market had virtually stopped funding companies that lacked large revenues outside the AI data centre, and predicted that more than 200 new clouds would experience mass collapse and integration, with only a few surviving. Hwang In-hoon believed that the funding problem could be resolved, while Mask believed that many enterprises might not be able to wait for that day.
On the one hand is the need for a sustained growth, and on the other is the growing focus on investment and debt-servicing pressures. Are the big chip buyers really running out of money?
The order went up fast and the cash didn't keep up.
Over the past few years, new clouds, such as CoreWeave and Lambda, have expanded their clusters through GPU mortgage loans, and funding providers have expanded from private lending institutions to large banks. Operators rely on equipment and customer contracts as a basis for financing, obtaining procurement funds in advance and then paying off their debts with hard rent. This arrangement supports rapid expansion and makes the next round of procurement increasingly dependent on how lenders assess the value of equipment and customer credit.
At the beginning of October, some of the Calculator Financing projects promoted by Weida required stronger guarantees from banks and regulators than had previously been the case. The lenders are concerned not only with whether the GPU can be rented today, but also with how much can be earned after a few years, how much can be paid by the customer and how much can be sold after default.
The judgement of GPU values has been divided between the parties. Tony Trzcinka, Senior Portfolio Manager at Impax Asset Management, stated that banks usually assessed GPU under a three- to four-year depreciation arrangement, while Inverda considered that a high-end GPU could generate up to 10 years of income. It is not the same thing that equipment can be used for long enough to generate sufficient income for it to continue. The lack of new clouds for other business income, more conservative mortgage valuations and higher collateral requirements may mean more self-financing or more expensive loans.
CoreWeave was also under pressure in the bond market. By mid-October, the market rate of return had risen to about 15 per cent as transaction prices fell. Investors require higher returns to take on their credit risk, and companies may also need to accept higher financing costs.
Its second quarter accounts showed a net cash flow from operating activities of approximately $679 million, while cash expenditure on the acquisition of equipment and property amounted to $6,422 million. During the same quarter, the net interest cost was approximately $640 million, equivalent to a quarter of the collection. Companies need to buy equipment for their customers' calculus, as well as interest on existing loans, and the continued expansion cannot be based solely on the current season.
On 20 October, CoreWeave responded to the debt doubt, emphasizing that most of the debt was supported by long-term customer contracts. The company obtains a commitment to purchase from its more creditworthy clients, on the basis of which it finances and builds its infrastructure; the contractual revenues from these contracts are sufficient to cover the underlying interest and do not have to rely on the customer for renewal. The corresponding income for the contracts to be completed at the end of the second quarter amounted to $104 billion, with an additional $25 billion in commitments at the beginning of the third quarter.
Contracts already exist to provide a source of repayment for the related obligations, but additional procurement still requires continued funding. If lenders require higher returns and more guarantees, the operator recalculates the cost-effectiveness of expanding the cluster, even if he receives new orders. The pressure on the new cloud further extends from the size of the input to the costs and conditions of each round of financing.
The sustainability of these procurement plans has also been included in the capacity assessment of the build-up. At the meeting on 15 October, the Wei Cheolers indicated that the company would consider such factors as cloud company capital expenditure, financial position, etc., and interact with end-users. The direct client of the Crystal Circle-generation plant is the chip designer, but advanced production and containment capacity needs to be pre-arranged, and the designer's demand will ultimately have to be supported by input from the end buyer. The verification of the financial position of these buyers will help to determine how long the demand for power is sustainable.
The chip manufacturers are also starting to raise money.
Weeda has supported several projects. The disclosure of transactional information on 28 October indicates that Zankore, Indonesia's Calculator Platform, is introducing more lending banks for $3.1 billion of GPU loans. It has a six-year agreement with Weeda, which includes a share of income and credit support, covering the unsold computing capacity of the first 100 MW facilities. Related financing had been announced in September, and recent disclosures had made it more specific how suppliers could be involved.
Following the general model described earlier by Weida, companies provide a minimum income commitment for part of their capacity to provide financial institutions with a basis to support project construction. New Clouds can still sell their calculations to third parties and, when the conditions are met, Young Wida participates in the income-sharing. Operators are supported beyond their own credit, while British Wida is given the opportunity to share the proceeds of subsequent leases in addition to hardware sales, with corresponding support responsibilities.
Model companies are also required to finance the equipment. At the end of the customized chip, Chase is already involved in the funding arrangements for Anthropic. On 26 October, Chase announced a $63 billion financing deal with its financial partners for the purchase by the new project company of its customized AI accelerator and network equipment for lease to Anthropic, which is expected to start in 2027. Banks and institutions such as Blackstone provide upfront purchase funds, and model companies obtain equipment by lease and access supplies according to schedule.
In order to establish this arrangement, Chase provided residual value support for the equipment. The regulatory document submitted on 27 October indicated that the maximum potential liability for all aircraft hangers, when deployed, would be approximately $43 billion, based on undefunct calibres. This increased liability as a result of random delivery, reduced by the payment of rent by Anthropic; if the customer defaults, Chase may bear the difference between the agreed outstanding amount and the disposal value of the equipment.
The $43 billion is the maximum amount of potential liability, not the cash already disbursed by Chase. The change is specific: after the supplier has completed the sale of the equipment, it will still have to look at whether the customer can pay the subsequent rent. The availability of chips and financing arrangements are further linked by funding from financial institutions, the client ' s phased cost of the equipment, and a share of the risk-taking by Chase in exchange for the procurement advance.
It supports the value of equipment, and in the above-mentioned model, it supports part of the revenue. Both approaches provide an additional layer of security for lenders, as well as an opportunity for some clients who cannot afford large amounts of inputs on their own to expand their procurement. For suppliers, the acquisition of such orders involves, in addition to product performance, delivery and price, the degree of support they are willing to provide to their clients.
But the lenders still have their own judgment. In early October, banking and asset management agencies involved in some of the associated financing projects in Weida required greater safeguards than the original programme. The involvement of the supplier did not eliminate the risk, but changed who took the risk. The continued expansion of the transaction depends on the financial institutions ' acceptance of these conditions and on the willingness of the chip manufacturer to reserve some subsequent responsibility for the new sales.
Funds are still in, but conditions vary.
On the second day of Mask ' s look at the new cloud financing, SpaceX announced that it would receive about $40 billion in GPU financing lease commitments. The entities managed by Apollo purchase GPU and associated equipment and lease it to SpaceX, and it is expected that the equipment will be delivered in batches in the first half of 2027. Financing still needs to meet the agreed terms of delivery, but the scale of commitments suggests that large GPU purchases can still attract financial institutions.
In explaining the financing advantages of SpaceX, Mask emphasized investment-grade credit, as well as revenues from the chain and launch operations. It is not the same as the new cloud that relies mainly on GPU rents to service its debt, and the source of funding that can be shown to lenders. There is no contradiction in looking at the financing prospects of some of the failing operators, while also financing their own expanded procurement; the difference is who is in a position to take on continuous inputs.
Cash from mature operations also allows greater scope for cloud manufacturers. Microsoft disclosed on 28 October that the quarterly cash flow of operating cash was $58 billion as at the end of September, that cash expenditure on the acquisition of equipment and property was $39.4 billion and that free cash flows remained at $18.6 billion. The company expects to spend approximately $170 billion on capital for the natural year 2026, and expects that the free cash flow will remain positive for the financial year 2027, while indicating that additional financing will be used when conditions are right.
Alphabet, previously mentioned, was financed by some $41.1 billion net through common shares and advanced unsecured instruments during the same quarter. Operating cash flows did not cover current capital inputs and did not prevent it from continuing to access capital markets. For such buyers, the raising of money is part of the arrangement of large amounts of inputs, and the ability to procure cannot be judged to be declining solely by the financing exercise.
New clouds are not always able to mobilize such business income, but some of them have already been supported by large client contracts. On 22 October, Nebius and Citigroup signed a $3.5 billion financing quota for data centre construction, GPU procurement, etc. The company indicated that long-term contracts of over $40 billion from investment-level clients could support the financing of subsequent assets, including Microsoft and Meta.
It also presents another layer of relationship between a cloud manufacturer and a new cloud. Microsoft and Meta invest in their own infrastructure and buy their power from external operators. New Clouds carries out hardware purchases and operations on behalf of clients, who in turn use their multi-year payment commitments for financing. The lenders see not only the size of the new cloud itself, but also who is on the other side of the lease and whether the contract income is reliable.
The Nebius financing period is 364 days and will be used in conjunction with client advances and operating cash flows to support the recent expansion. Long-term funding for 2027 and beyond will need to be continued. Large client contracts did not address all funding requirements at one time, but provided the conditions for a phased procurement and differentiated their fund-raising capacity from operators lacking stable tenants.
On 27 October, a three-year loan of approximately $129 million at a fixed annual interest rate of 9 per cent was also announced by the BUZA Alpha Cloud under the HIV/AIDS flag for the purchase of the GB200 cluster, which is planned to be repaid from the three-year contract income of Cohere. The loan covers approximately 80 per cent of the acquisition cost, with the remaining funding coming from previous reversible financing and the equipment is expected to be delivered by the end of the year.
Specific funding arrangements have been made for this procurement, as a result of the existence of a customer, a contract with a matching duration and a partial contribution from the operator. The 9 per cent fixed interest rate also serves as a reminder to the market that raising money is not the same as making it cheap. The new cloud needs to be compared not only with the GPU rental and equipment prices, but also with the borrowing costs and how they affect the scale of the next purchase.
Concluding remarks
Large buyers of chips are not all caught up in financial depletion, but their ability to raise funds is being tested by continuing to expand procurement. The mature business income, long-term customer contracts and the support provided by suppliers determine the conditions under which different buyers can continue to invest. In addition to the size of the demand for the chip industry, it is up to who can make the planned payments, complete the procurement and how long this input can last.
Auto-fetched and translated by AI Gem Box. All rights belong to the original author.