{"success":true,"data":{"pressRelease":{"id":"155534","rtpr_id":"nBw6v6cBSa-20260929","ticker":"QMLS","exchange":"NASDAQ","all_tickers":["QMLS"],"title":"QumulusAI Signs $240.9 Million in New AI Inference Customer Contracts for NVIDIA Blackwell B300 Capacity","author":"Business Wire","published_at":"2026-09-29T13:08:00.193Z","article_body":"QumulusAI Signs $240.9 Million in New AI Inference Customer Contracts for\nNVIDIA Blackwell B300 Capacity\n\nThe agreements cover more than 2,000 B300 GPUs and expand relationships first\nannounced on July 28 and July 23, 2026, taking the two customers' combined\ncontracted commitments with QumulusAI past $344 million. Extension options\nwould bring the new agreements' total potential value to $401.6 million over\nfive years\n\nQumulusAI (Nasdaq: QMLS), a neocloud infrastructure provider purpose-built for\nthe AI computing era, today announced two three-year take or pay agreements\nwith AI inference customers. The agreements have combined contracted value of\n$240.9 million and cover more than 2,000 NVIDIA Blackwell B300 GPUs across two\ndedicated clusters.\n\n$211.9 million, three years: an AI inference platform that helps companies\ndeploy large language models, vision models, speech models and other AI\napplications with low latency and high reliability. This is the customer's\nsecond agreement with QumulusAI. It follows the $71.9 million, three-year\nagreement announced July 28, 2026, and brings the customer's contracted\ncommitments with the company to more than $283 million. Capacity is targeted\nfor turnover to the customer in early 2027.\n\n$29.1 million, three years: a generative-AI inference platform serving\ndevelopers and enterprises running image, video and other generative models.\nThis is the customer's second agreement with QumulusAI. It follows the $32\nmillion, two-year agreement announced July 23, 2026, and brings the customer's\ncontracted commitments with the company to more than $61 million. Capacity is\nexpected to be turned over to the customer in December 2026.\n\n\"We're excited to keep growing alongside AI builders that are changing what\nthis industry can do, and it says a lot that both of these customers came back\nto commit to three more years with us,\" said Michael Maniscalco, CEO of\nQumulusAI. \"Inference platforms are planning capacity years out now, and they\nare backing those plans with cash. That lets us place GPUs against signed\ncontracts rather than against forecasts.\"\n\nConditions and qualifications. The $240.9 million of contracted value covers\nthe three-year initial terms only. The larger agreement carries two one-year\nextension options valued at $70.6 million each, which would bring its total\npotential value to $353.1 million over five years. The smaller agreement\ncarries a two-year extension option valued at $19.4 million, which would bring\nits total potential value to $48.4 million. The options may or may not be\nexercised, and their value is excluded from the contracted commitments stated\nabove.\n\nAbout QumulusAI\n\nQumulusAI is a distributed AI cloud platform that delivers accelerated access\nto high-performance GPU compute. Through an inference-first, demand-led\ndeployment model across a network of data center sites, QumulusAI brings\ncompute closer to customer demand, helping AI teams and enterprises scale\nproduction AI workloads with speed, flexibility and control. By combining\nrapid deployment with flexible private cloud infrastructure, QumulusAI gives\ncustomers a faster, more adaptable path beyond the capacity constraints of\ntraditional centralized and hyperscale cloud models. Learn more at\nQumulusAI.com\n(https://cts.businesswire.com/ct/CT?id=smartlink&url=https%3A%2F%2FQumulusAI.com&esheet=54612590&newsitemid=20260929733963&lan=en-US&anchor=QumulusAI.com&index=1&md5=06bbbf5e078b57bc6d56a625fa332bcc)\n.\n\nFollow us on LinkedIn (linkedin.com/company/qumulusai\n(https://cts.businesswire.com/ct/CT?id=smartlink&url=https%3A%2F%2Fwww.linkedin.com%2Fcompany%2Fqumulusai&esheet=54612590&newsitemid=20260929733963&lan=en-US&anchor=linkedin.com%2Fcompany%2Fqumulusai&index=2&md5=d07843312b6361bd92e216c589e290db)\n) and X @QumulusAI.\n\nForward-Looking Statements\n\nThis press release contains forward-looking statements within the meaning of\nSection 27A of the Securities Act of 1933, as amended, and Section 21E of the\nSecurities Exchange Act of 1934, as amended. These include statements\nregarding the company's agreements with two AI inference platform providers\nfor NVIDIA Blackwell B300 capacity; the anticipated value and term of the\nagreements; the potential exercise of the extension options and the total\npotential value of the agreements; the customers' total contracted commitments\nwith the company; the expected timing of capacity turnover under each\nagreement; and the company's continued growth. Words such as \"anticipate,\"\n\"believe,\" \"estimate,\" \"expect,\" \"intend,\" \"can,\" \"may,\" \"plan,\" \"project,\"\n\"target,\" \"will\" and similar expressions are intended to identify\nforward-looking statements. These statements are based on management's current\nexpectations and assumptions as of the date of this release. They are subject\nto risks and uncertainties that could cause actual results to differ\nmaterially, including, among others: the company's dependence on a limited\nnumber of large customers; delays in the delivery of GPUs and other equipment\nthe company has on order; the availability and cost of power, network\nconnectivity and specialized hardware such as graphics processing units; the\ncompany's substantial capital requirements and access to financing, including\nequipment lease financing; competition and rapid technological change in the\nhigh-performance computing and AI markets; the company's limited operating\nhistory and history of net losses; and those described in the \"Risk Factors\"\nsection of the company's registration statement on Form S-1, as amended (File\nNo. 333-292514), and the company’s quarterly report on Form 10-Q for the\nquarter ended June 30, 2026, as amended, filed with the U.S. Securities and\nExchange Commission (SEC), as such factors may be updated in the company's\nsubsequent filings with the SEC. QumulusAI undertakes no obligation to update\nor revise any forward-looking statement, whether as a result of new\ninformation, future developments or otherwise, except as required by\napplicable law.\n\n\n\nView source version on businesswire.com:\nhttps://www.businesswire.com/news/home/20260929733963/en/\n(https://www.businesswire.com/news/home/20260929733963/en/)\n\nInvestor Contact\ninvestors@qumulusai.com (mailto:investors@qumulusai.com)\n\nMedia Contact\nmedia@qumulusai.com (mailto:media@qumulusai.com)\n\n\nCopyright Business Wire 2026","article_body_html":"","raw_payload":{"data":{"id":"nBw6v6cBSa-20260929","title":"QumulusAI Signs $240.9 Million in New AI Inference Customer Contracts for NVIDIA Blackwell B300 Capacity","author":"Business Wire","ticker":"QMLS","created":"2026-09-29T13:08:00.193Z","tickers":["QMLS"],"exchange":"NASDAQ","article_body":"QumulusAI Signs $240.9 Million in New AI Inference Customer Contracts for\nNVIDIA Blackwell B300 Capacity\n\nThe agreements cover more than 2,000 B300 GPUs and expand relationships first\nannounced on July 28 and July 23, 2026, taking the two customers' combined\ncontracted commitments with QumulusAI past $344 million. Extension options\nwould bring the new agreements' total potential value to $401.6 million over\nfive years\n\nQumulusAI (Nasdaq: QMLS), a neocloud infrastructure provider purpose-built for\nthe AI computing era, today announced two three-year take or pay agreements\nwith AI inference customers. The agreements have combined contracted value of\n$240.9 million and cover more than 2,000 NVIDIA Blackwell B300 GPUs across two\ndedicated clusters.\n\n$211.9 million, three years: an AI inference platform that helps companies\ndeploy large language models, vision models, speech models and other AI\napplications with low latency and high reliability. This is the customer's\nsecond agreement with QumulusAI. It follows the $71.9 million, three-year\nagreement announced July 28, 2026, and brings the customer's contracted\ncommitments with the company to more than $283 million. Capacity is targeted\nfor turnover to the customer in early 2027.\n\n$29.1 million, three years: a generative-AI inference platform serving\ndevelopers and enterprises running image, video and other generative models.\nThis is the customer's second agreement with QumulusAI. It follows the $32\nmillion, two-year agreement announced July 23, 2026, and brings the customer's\ncontracted commitments with the company to more than $61 million. Capacity is\nexpected to be turned over to the customer in December 2026.\n\n\"We're excited to keep growing alongside AI builders that are changing what\nthis industry can do, and it says a lot that both of these customers came back\nto commit to three more years with us,\" said Michael Maniscalco, CEO of\nQumulusAI. \"Inference platforms are planning capacity years out now, and they\nare backing those plans with cash. That lets us place GPUs against signed\ncontracts rather than against forecasts.\"\n\nConditions and qualifications. The $240.9 million of contracted value covers\nthe three-year initial terms only. The larger agreement carries two one-year\nextension options valued at $70.6 million each, which would bring its total\npotential value to $353.1 million over five years. The smaller agreement\ncarries a two-year extension option valued at $19.4 million, which would bring\nits total potential value to $48.4 million. The options may or may not be\nexercised, and their value is excluded from the contracted commitments stated\nabove.\n\nAbout QumulusAI\n\nQumulusAI is a distributed AI cloud platform that delivers accelerated access\nto high-performance GPU compute. Through an inference-first, demand-led\ndeployment model across a network of data center sites, QumulusAI brings\ncompute closer to customer demand, helping AI teams and enterprises scale\nproduction AI workloads with speed, flexibility and control. By combining\nrapid deployment with flexible private cloud infrastructure, QumulusAI gives\ncustomers a faster, more adaptable path beyond the capacity constraints of\ntraditional centralized and hyperscale cloud models. Learn more at\nQumulusAI.com\n(https://cts.businesswire.com/ct/CT?id=smartlink&url=https%3A%2F%2FQumulusAI.com&esheet=54612590&newsitemid=20260929733963&lan=en-US&anchor=QumulusAI.com&index=1&md5=06bbbf5e078b57bc6d56a625fa332bcc)\n.\n\nFollow us on LinkedIn (linkedin.com/company/qumulusai\n(https://cts.businesswire.com/ct/CT?id=smartlink&url=https%3A%2F%2Fwww.linkedin.com%2Fcompany%2Fqumulusai&esheet=54612590&newsitemid=20260929733963&lan=en-US&anchor=linkedin.com%2Fcompany%2Fqumulusai&index=2&md5=d07843312b6361bd92e216c589e290db)\n) and X @QumulusAI.\n\nForward-Looking Statements\n\nThis press release contains forward-looking statements within the meaning of\nSection 27A of the Securities Act of 1933, as amended, and Section 21E of the\nSecurities Exchange Act of 1934, as amended. These include statements\nregarding the company's agreements with two AI inference platform providers\nfor NVIDIA Blackwell B300 capacity; the anticipated value and term of the\nagreements; the potential exercise of the extension options and the total\npotential value of the agreements; the customers' total contracted commitments\nwith the company; the expected timing of capacity turnover under each\nagreement; and the company's continued growth. Words such as \"anticipate,\"\n\"believe,\" \"estimate,\" \"expect,\" \"intend,\" \"can,\" \"may,\" \"plan,\" \"project,\"\n\"target,\" \"will\" and similar expressions are intended to identify\nforward-looking statements. These statements are based on management's current\nexpectations and assumptions as of the date of this release. They are subject\nto risks and uncertainties that could cause actual results to differ\nmaterially, including, among others: the company's dependence on a limited\nnumber of large customers; delays in the delivery of GPUs and other equipment\nthe company has on order; the availability and cost of power, network\nconnectivity and specialized hardware such as graphics processing units; the\ncompany's substantial capital requirements and access to financing, including\nequipment lease financing; competition and rapid technological change in the\nhigh-performance computing and AI markets; the company's limited operating\nhistory and history of net losses; and those described in the \"Risk Factors\"\nsection of the company's registration statement on Form S-1, as amended (File\nNo. 333-292514), and the company’s quarterly report on Form 10-Q for the\nquarter ended June 30, 2026, as amended, filed with the U.S. Securities and\nExchange Commission (SEC), as such factors may be updated in the company's\nsubsequent filings with the SEC. QumulusAI undertakes no obligation to update\nor revise any forward-looking statement, whether as a result of new\ninformation, future developments or otherwise, except as required by\napplicable law.\n\n\n\nView source version on businesswire.com:\nhttps://www.businesswire.com/news/home/20260929733963/en/\n(https://www.businesswire.com/news/home/20260929733963/en/)\n\nInvestor Contact\ninvestors@qumulusai.com (mailto:investors@qumulusai.com)\n\nMedia Contact\nmedia@qumulusai.com (mailto:media@qumulusai.com)\n\n\nCopyright Business Wire 2026"},"type":"article","timestamp":"2026-09-29T13:08:00.324997622Z","server_sent_at_ms":1790687280324},"received_at":"2026-09-29T13:08:00.381Z","source_url":"https://www.businesswire.com/news/home/20260929733963/en/"},"analysis":{"id":"144261","press_release_id":"155534","analysis_json":{"industry":{"label":"IT Services","sector":"Information Technology"},"redFlags":["customer concentration: limited number of large customers flagged in risk factors","company discloses limited operating history and history of net losses","extension option value ($160.7M of the $401.6M) may never be exercised","GPU delivery and power availability risk flagged by management"],"eventType":"operations_update","narrative":"QumulusAI signed two three-year take-or-pay AI inference contracts worth a combined $240.9 million, covering more than 2,000 NVIDIA Blackwell B300 GPUs across two dedicated clusters.\n\nBoth customers are repeat signers: the larger $211.9 million deal brings that customer's commitments past $283 million, and the $29.1 million deal pushes the other past $61 million — combined commitments now exceed $344 million.\n\nExtension options could lift the new agreements' total potential value to $401.6 million over five years. Capacity turns over in December 2026 and early 2027.","sentiment":"bullish","agentHooks":{"shouldPost":true,"suggestedAngle":"Repeat-customer take-or-pay contracts of $240.9M validate QumulusAI's contract-backed GPU deployment model ahead of datacenter capacity turnover."},"keyFigures":{"dealValueUsd":240900000,"customDimensions":{"gpus_covered":"more than 2,000","contract_term":"three years, take or pay","turnover_timing":"December 2026 (smaller contract) and early 2027 (larger contract)","larger_contract_value":"$211.9 million over three years","smaller_contract_value":"$29.1 million over three years","total_potential_value_usd":401600000,"larger_contract_extensions":"two one-year options at $70.6 million each, $353.1 million total potential","smaller_contract_extension":"two-year option at $19.4 million, $48.4 million total potential","combined_customer_commitments":"past $344 million"}},"quotedText":"That lets us place GPUs against signed contracts rather than against forecasts.","namedEntities":{"people":[{"name":"Michael Maniscalco","role":"CEO of QumulusAI"}],"products":["NVIDIA Blackwell B300 GPUs"],"companies":[{"name":"QumulusAI","ticker":"QMLS","relationship":"filer"},{"name":"AI inference platform customer","relationship":"customer ($211.9M three-year agreement)"},{"name":"generative-AI inference platform customer","relationship":"customer ($29.1M three-year agreement)"}],"dollarAmounts":[{"amount":"$240.9 million","context":"combined contracted value of the two new agreements"},{"amount":"$211.9 million","context":"larger three-year agreement with AI inference platform customer"},{"amount":"$29.1 million","context":"smaller three-year agreement with generative-AI inference platform"},{"amount":"$71.9 million","context":"prior three-year agreement announced July 28, 2026"},{"amount":"$32 million","context":"prior two-year agreement announced July 23, 2026"},{"amount":"$401.6 million","context":"total potential value of new agreements including extension options"},{"amount":"$344 million","context":"two customers' combined contracted commitments with QumulusAI"},{"amount":"$283 million","context":"larger customer's total contracted commitments"},{"amount":"$61 million","context":"smaller customer's total contracted commitments"},{"amount":"$70.6 million","context":"each one-year extension option on larger agreement"},{"amount":"$353.1 million","context":"five-year total potential value of larger agreement"},{"amount":"$19.4 million","context":"two-year extension option on smaller agreement"},{"amount":"$48.4 million","context":"five-year total potential value of smaller agreement"}]},"materialImpact":{"score":4,"reasoning":"Two new three-year take-or-pay contracts totaling $240.9 million ($401.6 million with extensions) for NVIDIA B300 capacity, with both customers returning for repeat commitments. Material revenue backlog for a neocloud provider, though delivery is spread over three years."},"tickerRelevance":{"others":[],"primary":"QMLS"},"globalImportance":45,"audienceRelevance":45,"eventTypeSecondary":["partnership"],"importanceComponents":{"tickerTier":"small-cap neocloud, high retail interest in AI infrastructure names","eventGravity":"large customer contract wins ($240.9M contracted, $401.6M potential)","retailFavoriteBoost":"AI/GPU exposure attracts retail attention","repeatCustomerSignal":"both customers renewing expands backlog credibility"}},"event_type":"operations_update","event_type_secondary":["partnership"],"sentiment":"bullish","material_impact_score":4,"narrative":"QumulusAI signed two three-year take-or-pay AI inference contracts worth a combined $240.9 million, covering more than 2,000 NVIDIA Blackwell B300 GPUs across two dedicated clusters.\n\nBoth customers are repeat signers: the larger $211.9 million deal brings that customer's commitments past $283 million, and the $29.1 million deal pushes the other past $61 million — combined commitments now exceed $344 million.\n\nExtension options could lift the new agreements' total potential value to $401.6 million over five years. Capacity turns over in December 2026 and early 2027.","key_figures":{"dealValueUsd":240900000,"customDimensions":{"gpus_covered":"more than 2,000","contract_term":"three years, take or pay","turnover_timing":"December 2026 (smaller contract) and early 2027 (larger contract)","larger_contract_value":"$211.9 million over three years","smaller_contract_value":"$29.1 million over three years","total_potential_value_usd":401600000,"larger_contract_extensions":"two one-year options at $70.6 million each, $353.1 million total potential","smaller_contract_extension":"two-year option at $19.4 million, $48.4 million total potential","combined_customer_commitments":"past $344 million"}},"named_entities":{"people":[{"name":"Michael Maniscalco","role":"CEO of QumulusAI"}],"products":["NVIDIA Blackwell B300 GPUs"],"companies":[{"name":"QumulusAI","ticker":"QMLS","relationship":"filer"},{"name":"AI inference platform customer","relationship":"customer ($211.9M three-year agreement)"},{"name":"generative-AI inference platform customer","relationship":"customer ($29.1M three-year agreement)"}],"dollarAmounts":[{"amount":"$240.9 million","context":"combined contracted value of the two new agreements"},{"amount":"$211.9 million","context":"larger three-year agreement with AI inference platform customer"},{"amount":"$29.1 million","context":"smaller three-year agreement with generative-AI inference platform"},{"amount":"$71.9 million","context":"prior three-year agreement announced July 28, 2026"},{"amount":"$32 million","context":"prior two-year agreement announced July 23, 2026"},{"amount":"$401.6 million","context":"total potential value of new agreements including extension options"},{"amount":"$344 million","context":"two customers' combined contracted commitments with QumulusAI"},{"amount":"$283 million","context":"larger customer's total contracted commitments"},{"amount":"$61 million","context":"smaller customer's total contracted commitments"},{"amount":"$70.6 million","context":"each one-year extension option on larger agreement"},{"amount":"$353.1 million","context":"five-year total potential value of larger agreement"},{"amount":"$19.4 million","context":"two-year extension option on smaller agreement"},{"amount":"$48.4 million","context":"five-year total potential value of smaller agreement"}]},"model_name":"glm-5.3-flashx","prompt_hash":"sha256:727b4b9429a443af","schema_hash":"sha256:05005c02d9cffac9","created_at":"2026-09-29T13:08:09.176Z","global_importance":45,"audience_relevance":45,"importance_components":{"tickerTier":"small-cap neocloud, high retail interest in AI infrastructure names","eventGravity":"large customer contract wins ($240.9M contracted, $401.6M potential)","retailFavoriteBoost":"AI/GPU exposure attracts retail attention","repeatCustomerSignal":"both customers renewing expands backlog credibility"}},"durationMs":8756,"modelName":"glm-5.3-flashx"}}