We have disclosed under the heading "Risk Factors" in our Annual Report the risk factors which materially affect our business, financial condition or results of operations. Other than as set forth below, there have been no material changes from the risk factors previously disclosed. You should carefully consider the risk factors set forth in our Annual Report and the other information set forth elsewhere in this Quarterly Report on Form 10-Q. You should be aware that these risk factors and other information may not describe every risk facing our company. Additional risks and uncertainties not currently known to us or that we currently deem to be immaterial also may materially adversely affect our business, financial condition and/or operating results. We depend on AMD as well as other Key Partners for our Enterprise AI business. Our Enterprise AI business depends substantially on our relationship with Advanced Micro Devices, Inc. ("AMD") as well as on other partners who provide, or may provide in the future, complementary products, software and platform capabilities for our Enterprise AI solutions, including Palantir Technologies, Inc., VMware by Broadcom, Uniphore Technologies, Inc., Rubrik, Inc. and any future partners (collectively, our "Key Partners"). Our ability to obtain AMD Products and related hardware on commercially acceptable terms and timelines depends on our relationships with the third-party original equipment manufacturers who incorporate AMD Products into their servers as well as supply and demand for GPUs and related hardware in the overall market, which can fluctuate rapidly. We have entered into a definitive GPU-as-a-Service Agreement - Master Terms and Conditions (the "GPUaaS Agreement") with AMD intended to facilitate the phased deployment of AMD AI compute products (including AMD Instinct™ GPUs (e.g., MI355X, MI350P, and future successor chips) and AMD EPYC™ CPUs) (collectively, the "AMD Products") in our global data centers. The GPUaaS Agreement establishes a commercial framework and does not itself obligate AMD to purchase any particular quantity of services, approve any deployment, or enter into future commercial arrangements. We have agreed to dedicate, maintain and make available an aggregate of 30 megawatts of capacity within our data centers as an initial launch footprint for AMD Products, where AMD Products are fit for purpose and subject to certain financing, operational, and legal conditions (the "AMD Footprint Obligation"). Satisfying the AMD Footprint Obligation will require significant capital expenditures, financing, customer demand, power availability, cooling capacity, networking infrastructure, operational execution and other conditions that may not occur on the anticipated timeline, at the anticipated costs, or at all. If we fail to satisfy the AMD Footprint Obligation, we could lose certain benefits under the GPUaaS Agreement, and our relationship with AMD could be adversely affected. Each deployment under the GPUaaS Agreement is subject to agreement between the parties on the commercial terms applicable to such deployment, including pricing, term and financial parameters, and AMD has no obligation under the GPUaaS Agreement to agree to any particular deployment as being within the scope of the framework of the GPUaaS Agreement. If the parties agree that a deployment is within the scope of the GPUaaS Agreement, AMD has agreed to purchase residual unsold capacity from the applicable deployment, subject to the satisfaction of delivery and service availability requirements and an aggregate cap for each deployment. There can be no assurance that the applicable conditions will be satisfied or that such purchases will offset our investment, financing or operating costs. The GPUaaS Agreement also provides AMD with a right of first refusal to consume GPU capacity before the Company sells GPUaaS services to third parties below a specified price threshold. The GPUaaS Agreement remains in effect until the expiration or termination of all deployments thereunder. Either party may terminate an affected deployment upon the other party's uncured material breach or insolvency, if the go-live date for such deployment has not occurred by a specified outside date, or upon certain service level failures. - 51 - Table of Contents We also depend on joint marketing and demand generation activities conducted together with AMD and our Key Partners to build customer awareness and generate sales opportunities for our Enterprise AI solutions. Any reduction in, or failure to effectively execute, these efforts could adversely affect demand for our Enterprise AI solutions and our results of operations. Demand for our Enterprise AI solutions may also be affected by customer preferences for GPU technology that may not align with the technology we have deployed or plan to deploy. AMD or our Key Partners may change product roadmaps, allocation priorities, commercialization strategies, pricing or licensing terms, interoperability, preferred deployment partners or channel relationships at any time, which could reduce our access to the products and services offered by AMD and our Key Partners or adversely affect the pricing, economics, or competitiveness of our Enterprise AI offerings. Our ability to realize anticipated benefits from these relationships depends upon successful integration of the platforms of our Key Partners into our offerings, customer adoption, continued commercial cooperation, and the ability of the parties to execute their obligations. Customers may elect not to adopt offerings provided by our Key Partners or may prefer competing platforms. If our relationship with AMD or any of our Key Partners deteriorates for any reason, our ability to deliver Enterprise AI solutions to customers, and our results of operations, could be materially and adversely affected. The debt we have incurred or may incur to deploy AI infrastructure could adversely affect our financial condition and results of operations. To fund the acquisition of GPU hardware and related AI infrastructure, we expect to incur substantial additional indebtedness, whether through equipment financing/leasing arrangements, secured term loans, vendor or OEM financing, or other instruments. Additional leverage will increase our overall debt service obligations, may impose restrictive covenants limiting our operational and financial flexibility, and increase our exposure to interest rate fluctuations and refinancing risk. Certain financing arrangements may require us to pledge GPUs or other assets as collateral, which could limit our financial flexibility. Our ability to service this debt depends on generating sufficient Enterprise AI revenue and infrastructure utilization to justify the underlying capital expenditure; if customer demand does not develop as anticipated, if pricing for AI compute declines due to competition or oversupply, or if we are not able to attach value-added services, we may be unable to generate returns sufficient to service this debt or recover our capital investment. In addition, there can be no assurances that we will be able to obtain financing on commercially reasonable terms or in the amounts necessary to facilitate such deployments, if at all. If we are unable to access capital markets or financing on acceptable terms in the future, our ability to expand or maintain our Enterprise AI business could be impaired. Our Enterprise AI business is capital-intensive and depends on recovering large investments in rapidly evolving GPU infrastructure, and if GPU utilization, customer demand, hardware residual values or other assumptions fall short of our expectations, our results of operations could be materially and adversely affected. Our Enterprise AI business is capital-intensive, and its profitability depends on maintaining sufficient utilization of deployed GPU infrastructure. Our expectations regarding the size, timing and growth of the Enterprise AI market, customer demand, utilization rates, pricing, financing availability, returns on invested capital and our ability to attach value-added services may prove inaccurate. Because a substantial portion of our operating and financing costs is fixed and continues to be incurred regardless of utilization, periods of underutilization, including if demand develops more slowly than we anticipate, or if we invest ahead of customer adoption, could materially reduce our margins and cash flows. The Enterprise AI infrastructure market is characterized by rapid technological change, and the GPUs and related infrastructure we acquire may become obsolete or non-competitive before we have recovered our investment, requiring further capital expenditure to remain competitive. The residual value of AI hardware is uncertain. If the market value of our GPU infrastructure declines more rapidly than expected, we could incur impairment charges, refinancing challenges or losses upon disposition. We may also face customer concentration risk if a limited number of large customers account for a significant portion of Enterprise AI revenue, and the loss of any such customer could materially affect our results. Any of these factors, individually or together, could prevent us from realizing an adequate return on our GPU infrastructure investments and could have a material adverse effect on our results of operations. - 52 - Table of Contents