D-Wave Quantum Inc. (QBTS) 2026 Q3 法說會逐字稿

Good morning, everyone. Welcome to D-Wave's third quarter fiscal year 2026 earnings conference call. Today's conference call is being recorded.

各位早安。歡迎參加 D-Wave 2026 會計年度第三季財報電話會議。今天的電話會議將進行錄音。

At this time, I would like to turn the call over to Kevin Hunt, Senior Director of Investor Relations. Please go ahead.

此刻，我想把電話會議交給投資人關係資深總監 Kevin Hunt。請開始。

Thank you, and good morning. With me today are Dr. Alan Baratz, our Chief Executive Officer; and Greg Golkov, our Acting Chief Financial Officer.

謝謝，各位早安。今天與我一同出席的有我們的執行長 Alan Baratz 博士，以及代理財務長 Greg Golkov。

Before we begin, I would like to remind everyone that this call may contain forward-looking statements and should be considered in conjunction with cautionary statements contained in our earnings release and the company's most recent periodic SEC reports. During today's call, management will provide certain information that will constitute non-GAAP financial measures under SEC rules, such as non-GAAP gross profit, non-GAAP gross margin, non-GAAP adjusted operating expenses, adjusted EBITDA loss, and operating metrics such as bookings. Reconciliations to GAAP financial measures and certain additional information are also included in today's earnings release, which is available in the Investor Relations section of our company website at www.dwavequantum.com.

在開始之前，我想提醒各位，本次電話會議可能包含前瞻性陳述，並應結合我們的財報新聞稿以及公司最近一期向美國證券交易委員會（SEC）提交的定期報告中所載的警示性陳述一併考量。在今天的電話會議中，管理團隊將提供若干依 SEC 規則構成非 GAAP 財務衡量指標的資訊，例如非 GAAP 毛利、非 GAAP 毛利率、非 GAAP 調整後營業費用、調整後 EBITDA 虧損，以及如訂單額（bookings）等營運指標。與 GAAP 財務衡量指標的調節表及若干補充資訊亦包含於今天的財報新聞稿中，該新聞稿可於公司網站 www.dwavequantum.com 的投資人關係（Investor Relations）專區取得。

I'll now hand over the call to Alan.

我現在把電話會議交給 Alan。

Thank you, Kevin. Good morning, and thank you all for joining us. I'm pleased to share with you today our third-quarter 2026 business results and highlights.

謝謝你，Kevin。各位早安，感謝各位加入我們。我很高興今天與各位分享我們 2026 年第三季的營運成果與重點。

Our third-quarter revenue was $4.5 million, up 21% year over year. Bookings for the first nine months of 2026 were $37.4 million, a 611% year-over-year increase. That triple-digit, year-over-year bookings growth and a robust pipeline reinforce our confidence in the opportunity ahead.

我們第三季營收為 450 萬美元，年增 21%。2026 年前九個月的訂單額（bookings）為 3,740 萬美元，年增 611%。這樣的三位數年增訂單成長以及強勁的商機管線，強化了我們對未來機會的信心。

We are expanding our commercial reach, delivering measurable customer value and advancing both our annealing and gate model roadmaps. As a reminder, D-Wave is the only dual-platform quantum computing company, and that gives us a distinctive position in this market.

我們正在擴大商業觸及範圍，交付可衡量的客戶價值，並推進我們在退火（annealing）與閘模型（gate model）兩條路線圖上的進展。提醒各位，D-Wave 是唯一的雙平台量子運算公司，這使我們在此市場中具備獨特定位。

Our annealing quantum computers are available today with customers using hybrid quantum applications in their operations. At the same time, we are developing fault-tolerant gate model systems designed to address a different set of computational challenges.

我們的退火量子電腦目前已可供使用，客戶正將混合式量子應用導入其營運。同時，我們也在開發容錯的閘模型系統，旨在解決另一類計算挑戰。

We are building our business around both opportunities. Let me start with our commercial progress. During the quarter, NTT DOCOMO, Japan's largest mobile operator, deployed its second production application powered by D-Wave's hybrid quantum technology. In addition, we continue to make progress with other large enterprise customers that are pursuing production applications, including UnitedHealth Group's Optum and AT&T.

我們的業務將圍繞這兩項機會來打造。我先從我們的商業進展談起。本季期間，日本最大行動電信業者 NTT DOCOMO 部署了第二個由 D-Wave 混合式量子技術驅動的正式生產應用。此外，我們也持續在其他正推動生產級應用的大型企業客戶上取得進展，包括 UnitedHealth Group 旗下的 Optum 與 AT&T。

We are also continuing to pursue system sales. In October, we delivered Italy's first annealing quantum computer, a D-Wave Advantage2 system for installation at Villa Saporiti in Como. Funded by Swiss Quantum Technology, the system will support Italy's efforts to accelerate application development, build quantum workforce skills, and advance commercial adoption.

我們也持續推進系統銷售。10 月，我們交付了義大利第一台退火量子電腦——D-Wave Advantage2 系統，將安裝於科莫（Como）的 Villa Saporiti。該系統由 Swiss Quantum Technology 出資，將支援義大利加速應用開發、培育量子人才技能，並推動商業採用。

System placements can create a focal point for a broader quantum ecosystem. Local access brings researchers, universities, and businesses together around the technology, creating opportunities to develop applications and build demand. System sales also allow us to serve organizations with different access requirements and strategic priorities.

系統部署可成為更廣泛量子生態系的聚焦點。在地化的存取能讓研究人員、大學與企業圍繞該技術凝聚在一起，創造開發應用並建立需求的機會。系統銷售也讓我們能服務具有不同存取需求與策略優先順序的組織。

Moving on to partners. Partners are beginning to play a stronger role in advancing our go-to-market efforts for both system sales and quantum compute as a service, as well as helping us move more customer applications into production deployments. Recent relationships are expanding our ability to deliver on that.

接著談合作夥伴。合作夥伴正開始在推進我們的市場進入（go-to-market）工作上扮演更重要的角色，涵蓋系統銷售與量子運算即服務（quantum compute as a service），並協助我們將更多客戶應用推進到生產部署。近期建立的合作關係正在擴大我們交付上述成果的能力。

CGI, one of the world's largest independent IT and business consulting services firms, plans to integrate D-Wave's quantum computing platforms into its advanced technology services portfolio. Following a year of collaboration, we expect this partnership will help clients develop and deploy quantum and hybrid quantum applications in industries including logistics, transportation, and retail.

CGI 是全球最大的獨立 IT 與商業顧問服務公司之一，計畫將 D-Wave 的量子運算平台整合至其先進技術服務產品組合中。在合作一年之後，我們預期此夥伴關係將協助客戶在包含物流、運輸與零售等產業中，開發並部署量子與混合式量子應用。

At our sold-out Qubits Asia User Conference in Seoul two weeks ago, we announced an MOU with LG CNS, a leading AI transformation company and a technology arm of LG Group. The collaboration will focus on enterprise optimization within LG Group and among LG CNS's customers in South Korea and beyond. LG CNS brings mathematical optimization capabilities, industry knowledge, and systems integration experience that complement our annealing quantum computing technology.

在兩週前於首爾舉行、門票售罄的 Qubits Asia 使用者大會上，我們宣布與 LG CNS（領先的 AI 轉型公司、亦為 LG 集團的技術部門）簽署合作備忘錄（MOU）。此合作將聚焦於 LG 集團內部以及 LG CNS 在南韓及其他地區客戶的企業最佳化。LG CNS 具備數學最佳化能力、產業知識與系統整合經驗，可與我們的退火量子運算技術形成互補。

These relationships address a practical requirement for accelerating quantum computing adoption. Enterprises need help identifying the right problems, developing applications, and integrating them into existing operations. Working with partners that already understand those customers and their systems can help us expand our reach and support deployment.

這些合作關係回應了加速量子運算採用的一項實務需求。企業需要協助來辨識合適的問題、開發應用，並將其整合進既有營運。與已了解這些客戶及其系統的合作夥伴合作，能幫助我們擴大觸及並支援部署。

As we expand our commercial reach through these partnerships, we are also strengthening the manufacturing relationships that support our technology roadmap. We recently extended our longstanding relationship with SkyWater through a multi-year agreement for semiconductor process development and manufacturing. That relationship also provides an opportunity to apply our technology within semiconductor manufacturing.

在透過這些合作夥伴擴大商業觸及的同時，我們也在強化支撐技術路線圖的製造合作關係。我們近期透過一項多年期協議，延長了與 SkyWater 長期以來在半導體製程開發與製造方面的合作。該合作關係也提供了在半導體製造中應用我們技術的機會。

A proof-of-technology project using D-Wave's annealing quantum computing technology to optimize maintenance schedules at SkyWater's fab reduced lot delays by up to 30%, improving factory throughput. That project will now advance to a proof of concept, incorporating resource availability, technician constraints, and additional fab-specific operational constraints.

一項技術驗證（proof-of-technology）專案使用 D-Wave 的退火量子運算技術來最佳化 SkyWater 晶圓廠的維護排程，使批次（lot）延遲最多降低 30%，提升了工廠產出效率。該專案現在將推進至概念驗證（proof of concept），納入資源可用性、技師限制，以及其他晶圓廠特定的營運限制。

SkyWater has been a key partner of D-Wave for more than 10 years, and we view them as a key partner going forward. That said, prudent supply chain management requires diversification of suppliers, so we will continue to qualify additional manufacturing sources.

SkyWater 作為 D-Wave 的關鍵合作夥伴已超過 10 年，我們也將其視為未來的重要合作夥伴。不過，審慎的供應鏈管理需要供應商多元化，因此我們將持續認證其他製造來源。

Now let me turn to our products and technology. Last week, we announced performance gains to our Stride hybrid solver, available through the Leap cloud service. For example, on assembly line balancing benchmark problems, the enhanced solver found better solutions in 15 seconds than the previous version found in 120 seconds.

接下來我談談我們的產品與技術。上週，我們宣布對 Stride 混合式求解器的效能提升，該求解器可透過 Leap 雲端服務使用。例如，在裝配線平衡（assembly line balancing）的基準測試問題上，強化後的求解器在 15 秒內找到的解，比先前版本在 120 秒內找到的解更佳。

For businesses making time-sensitive decisions, both solution quality and speed matter. A better production schedule or resource allocation is most useful when it is available in time to act. These enhancements are designed to help our customers obtain better answers within the time their businesses require.

對於需要做出時間敏感決策的企業而言，解的品質與速度同等重要。更好的生產排程或資源配置，只有在能及時提供以便採取行動時，才最有價值。這些強化旨在協助客戶在其業務所需的時間內取得更好的答案。

We are also expanding the capabilities of our annealing quantum computers. Earlier this year, we published a preprint demonstrating analog-digital quantum computation on our annealing quantum computers, extending the range of quantum dynamics we can investigate. Since annealing dynamics are inherently analog, like the dynamics of quantum materials, we can avoid much of the overhead that gate model systems require to simulate quantum materials. And unlike in lab experiments, we can investigate the behavior of every individual spin and program material properties without synthesizing new samples.

我們也在擴展退火量子電腦的能力。今年稍早，我們發表了一篇預印本（preprint），展示在我們的退火量子電腦上進行類比—數位量子計算（analog-digital quantum computation），擴大了我們可研究的量子動力學範圍。由於退火動力學本質上是類比的，類似於量子材料的動力學，我們可以避免閘模型系統在模擬量子材料時所需的大量額外負擔。而且不同於實驗室實驗，我們可以研究每一個單獨自旋（spin）的行為，並在不合成新樣品的情況下對材料性質進行編程。

In October, we published a preprint describing experiments realizing topological insulators in one, two, and three dimensions using more than 700 qubits. We believe these include the first programmable quantum dynamics experiments in three dimensions. This work advances our ability to investigate the behavior of quantum materials and broadens the scientific opportunity for our annealing systems.

10月，我們發表了一篇預印本，描述使用超過700個量子位元在一維、二維與三維中實現拓撲絕緣體的實驗。我們相信其中包含首批可程式化的三維量子動力學實驗。這項工作提升了我們研究量子材料行為的能力，並擴大了我們退火系統的科學研究機會。

On our annealing hardware roadmap, we are advancing two technologies central to scaling. That includes superconducting interconnect that enables multi-chip systems and scalable I/O control circuitry that supports significantly larger processors. We have already shown experimentally that we can achieve strong coupling between prototype annealing fabrics that are interconnected across two chips.

在我們的退火硬體路線圖上，我們正推進兩項對擴展規模至關重要的技術。其中包括可實現多晶片系統的超導互連，以及支援顯著更大處理器的可擴展I/O控制電路。我們已在實驗上展示，能在跨兩片晶片互連的原型退火晶格之間達成強耦合。

Building on this foundational work, we have completed design of a dual-chip annealing processor and are releasing it for fabrication this week. In addition, we are on track to complete fabrication of our first scalable I/O development wafers in November. We believe these are important engineering steps toward the larger systems on our roadmap, including 20,000-qubit and ultimately 100,000-qubit annealing systems.

在這些基礎工作之上，我們已完成雙晶片退火處理器的設計，並將於本週送交製造。此外，我們也按計畫將於11月完成首批可擴展I/O開發晶圓的製造。我們相信，這些是邁向路線圖上更大型系統的重要工程步驟，包括20,000量子位元，並最終達到100,000量子位元的退火系統。

Turning to gate model development, our strategy starts with the requirements for useful computation -- logical qubits that can execute sufficiently long calculations reliably with practical hardware requirements and execution times. Our differentiated dual-rail architecture features built-in error detection, which is designed to reduce the hardware overhead required for quantum error correction.

轉向閘模型開發，我們的策略從實用計算的需求出發——能以可行的硬體需求與執行時間，可靠地執行足夠長計算的邏輯量子位元。我們具差異化的雙軌架構具備內建錯誤偵測，旨在降低量子錯誤更正所需的硬體額外負擔。

Combined with the speed of superconducting operations, this underpins our approach to building scalable, fault-tolerant systems. To prepare customers and researchers to use our forthcoming gate model systems, we launched a simulator beta program in early October. Participants include both commercial and research organizations, including BBVA, FirstQFM, Florida Atlantic University, and the Jülich Supercomputing Centre. Built around our dual-rail superconducting technology, the simulator will enable users to experiment with error-aware programming and develop quantum applications in advance of hardware availability.

再結合超導操作的速度，這奠定了我們打造可擴展、容錯系統的方法。為了讓客戶與研究人員能使用我們即將推出的閘模型系統，我們在10月初啟動了模擬器Beta計畫。參與者包括商業與研究機構，例如BBVA、FirstQFM、佛羅里達大西洋大學，以及于利希超級計算中心（Jülich Supercomputing Centre）。該模擬器以我們的雙軌超導技術為核心，將使使用者能進行具錯誤意識的程式設計實驗，並在硬體可用之前提前開發量子應用。

As we have previously published on our gate model roadmap, we are on track to complete a 17-qubit gate model system, and we are already discussing 17-qubit system orders for delivery in 2027. Our longer-term roadmap targets 100 logical qubits capable of reliably executing more than 1 million operations by 2032.

如同我們先前在閘模型路線圖中所發布的內容，我們正按計畫完成一套17量子位元的閘模型系統，並且已開始討論於2027年交付的17量子位元系統訂單。我們較長期的路線圖目標是在2032年前達成100個邏輯量子位元，能可靠地執行超過100萬次操作。

Now let me make some clarifying comments on logical qubits, as there seem to be different definitions of logical qubits floating around, some of them a bit misleading. First, at D-Wave, we are focused on solving real-world customer problems, so our roadmap is focused on metrics that define a useful system. It is generally accepted that you need a system with at least 100 fault-tolerant logical qubits to solve a useful real-world problem with fault tolerance defined as the ability to perform at least 1 million gate operations successfully.

現在我想就邏輯量子位元做一些釐清說明，因為目前似乎存在不同的邏輯量子位元定義，其中有些略具誤導性。首先，在D-Wave，我們專注於解決客戶的真實世界問題，因此我們的路線圖聚焦於界定「有用系統」的衡量指標。一般而言，普遍接受的看法是：要在具容錯能力的情況下解決有用的真實世界問題，你需要至少100個容錯邏輯量子位元；而此處的容錯定義為能成功執行至少100萬次閘操作的能力。

Second, as of today, we are not aware of any quantum computing company that has produced even one logical qubit that meets that fault-tolerant standard. Our gate model roadmap lays out a timeline to achieve the first fully error-corrected, fault-tolerant logical qubit by 2028 and then 100 by 2032. Third, many of our peers are using alternative definitions for logical qubits, some only error detected or partially error corrected or at error rates below the generally accepted fault-tolerant threshold.

第二，截至今日，我們尚未得知有任何量子運算公司已產生哪怕一個符合該容錯標準的邏輯量子位元。我們的閘模型路線圖規劃在2028年前達成第一個完全錯誤更正、具容錯能力的邏輯量子位元，並在2032年前達到100個。第三，我們許多同業採用替代性的邏輯量子位元定義，有些僅做到錯誤偵測或部分錯誤更正，或是在錯誤率低於一般公認的容錯門檻之下。

So the point is, when assessing vendors' roadmaps, it is important to clarify and fully understand the terms and definitions that underlie those roadmaps. D-Wave is moving forward with an approach to error correction that is designed to be efficient, scalable and faster. And that's why we believe we'll be the first to deliver a scalable, fault-tolerant gate model system.

因此重點是，在評估供應商的路線圖時，釐清並充分理解支撐這些路線圖的術語與定義非常重要。D-Wave正推進一種旨在更高效率、可擴展且更快速的錯誤更正方法。這也是我們相信自己將率先交付可擴展、具容錯能力的閘模型系統的原因。

Now let me turn to government, an area with some significant announcements over the past few months. We executed a definitive agreement with the US Department of Commerce, following our previously announced letter of intent, providing access to up to $100 million of US CHIPS and Science Act funding. The funding is intended to support development of our next-generation annealing and gate model quantum computers. These include the 100,000-qubit annealing quantum computing, targeting greater performance across optimization, materials simulation, blockchain, and artificial intelligence, as well as the 10,000-qubit gate model system designed to enable 100 logical qubits capable of reliably executing more than 1 million operations. We believe this agreement also reflects US government validation of both our annealing and gate model roadmaps.

接下來我談談政府領域，過去幾個月在此方面有一些重要公告。在先前公布意向書之後，我們已與美國商務部簽署最終協議，可取得最高達1億美元的《美國CHIPS與科學法案》資金。該資金旨在支持我們下一代退火與閘模型量子電腦的研發。其中包括100,000量子位元的退火量子運算系統，目標是在最佳化、材料模擬、區塊鏈與人工智慧等方面提供更高效能；以及10,000量子位元的閘模型系統，旨在實現100個邏輯量子位元，能可靠地執行超過100萬次操作。我們相信，這項協議也反映了美國政府對我們退火與閘模型路線圖的肯定。

We are also seeing recognition extend to third-party industry analyst firms. If you recall, last quarter, IDC named D-Wave as one of only two companies in the leaders category in their 2026 quantum computing vendor assessment. Recently, Gartner recognized D-Wave as a Market Shaper in its Emerging Market Quadrant for Quantum Computing Established Vendors as of November 2026. And we were one of only six vendors placed in the upper right quadrant.

我們也看到第三方產業分析機構的認可持續擴大。如果各位還記得，上季IDC在其2026年量子運算供應商評估中，將D-Wave列為領導者類別中僅有的兩家公司之一。近期，Gartner在其截至2026年11月的《量子運算既有供應商》新興市場象限中，將D-Wave評為市場塑造者（Market Shaper）。而我們也是僅有六家被放入右上象限的供應商之一。

Before I turn the call over to Greg, I want to recognize several additions to our Board of Directors. Bernard Gavgani, former group CIO of BNP Paribas, joined our Board and Cybersecurity Committee, bringing deep enterprise technology experience. Kevan Krysler joined the Board and Audit committee, adding extensive expertise in public company finance, corporate growth strategy, and risk management.

在我把電話會議交給Greg之前，我想介紹我們董事會的幾位新成員。法國巴黎銀行（BNP Paribas）前集團資訊長Bernard Gavgani加入我們董事會與資安委員會，帶來深厚的企業技術經驗。Kevan Krysler加入董事會與審計委員會，增添在上市公司財務、企業成長策略與風險管理方面的豐富專業。

So in summary, we are focused on expanding customer adoption, converting our pipeline into bookings and revenue, and delivering on our product roadmap. Our commercial results and technical progress reinforce our confidence in the opportunity ahead.

總結來說，我們專注於擴大客戶採用、將我們的商機管線轉化為訂單與營收，並落實我們的產品路線圖。我們的商業成果與技術進展，強化了我們對未來機會的信心。

With that, I will hand the call over to Greg to provide a review of our third-quarter results. Greg?

接下來，我將把電話會議交給Greg，請他回顧我們第三季的業績。Greg？

Thank you, Alan, and thank you to everyone for taking the time to participate in our third-quarter earnings call. Revenue in the third quarter of 2026 totaled $4.5 million, an increase of approximately 21% year over year. In the third quarter, D-Wave recognized revenue from more than 100 customers and continued its strong momentum with commercial enterprises, with these customers accounting for 52% of total revenue in the quarter, up more than 20 percentage points from 31% in the third quarter of 2025.

謝謝你，Alan，也謝謝各位撥冗參加我們第三季財報電話會議。2026年第三季營收合計為450萬美元，年增約21%。在第三季，D-Wave來自超過100位客戶的營收入帳，並在商業企業客戶方面延續強勁動能；這些客戶在本季占總營收的52%，較2025年第三季的31%提高了超過20個百分點。

From a product perspective, Q3 revenue was comprised of the following: $2.2 million in QCaaS subscription revenue, up 56% year over year; $2 million in professional services revenue, up 318% year over year, driven primarily by a significant engagement with a Fortune 100 customer and paid research projects related to gate model quantum computing; and approximately $360,000 in systems and other revenue. Bookings for the third quarter totaled $1.9 million, a year-over-year decrease of $435,000 or 19%.

從產品角度來看，第三季營收由以下項目構成：QCaaS訂閱營收220萬美元，年增56%；專業服務營收200萬美元，年增318%，主要由與一家《財富》100大客戶的重要合作案，以及與閘模型量子運算相關的付費研究專案所帶動；以及約36萬美元的系統與其他營收。第三季訂單金額（bookings）合計為190萬美元，較去年同期減少43.5萬美元，或19%。

Note that bookings in any single quarter can vary significantly based on the timing of larger transactions, and year-to-date bookings for the first nine months of the year were 37.4 million, a 611% year over year increase. Bookings for the first nine months included a $20 million system sale to Florida Atlantic University, or FAU. We expect to deliver that system and recognize a substantial portion of the related revenue in the fourth quarter.

請注意，任何單一季度的訂單金額可能會因較大型交易的時點而大幅波動，而本年度截至目前（年初至今）前九個月的訂單金額為 3,740 萬美元，較去年同期成長 611%。前九個月的訂單包含對佛羅里達大西洋大學（Florida Atlantic University，簡稱 FAU）的一筆 2,000 萬美元系統銷售。我們預期將於第四季交付該系統，並在第四季認列相關收入中的相當大一部分。

Our sales pipeline continued to expand in the third quarter as dollar value increased more than 75% from the end of the second quarter and has more than tripled since the end of 2025, driven by strong interest in system sales. As of September 30, 2026, the aggregate amount of remaining performance obligations, or RPO, totaled $38.3 million, more than 13 times the September 30, 2025, RPO balance. 59% of this quarter's RPO balance is expected to be recognized as revenue within the following 12 months, and 71% is expected to be recognized within the following 24 months.

我們的銷售管線在第三季持續擴大，美元價值較第二季末增加逾 75%，且自 2025 年底以來已增加至三倍以上，主要受系統銷售的強勁需求所帶動。截至 2026 年 9 月 30 日，剩餘履約義務（remaining performance obligations，RPO）總額為 3,830 萬美元，較 2025 年 9 月 30 日的 RPO 餘額增加逾 13 倍。本季 RPO 餘額中，預計有 59% 將於未來 12 個月內認列為收入，預計有 71% 將於未來 24 個月內認列為收入。

Turning to gross profit and margin, on a GAAP basis, gross profit for the third quarter of 2026 was $2.4 million, a year-over-year decrease of $0.3 million or 11%, while gross margin was 52.9%, a year-over-year decrease of 18.5%. Non-GAAP gross profit for the quarter was $2.9 million, essentially flat year over year. And non-GAAP gross margin was 63.6%, a decrease of 14.1% year over year.

接著談毛利與毛利率。按 GAAP 基礎，2026 年第三季毛利為 240 萬美元，較去年同期減少 30 萬美元或 11%，毛利率為 52.9%，較去年同期下降 18.5%。本季非 GAAP 毛利為 290 萬美元，與去年同期基本持平。本季非 GAAP 毛利率為 63.6%，較去年同期下降 14.1%。

Gross margin can fluctuate from quarter to quarter and vary significantly depending on product mix. The decrease in gross profit and gross margin this quarter was primarily due to a higher proportion of professional services revenue that carries a lower margin.

毛利率可能在各季度間波動，並會因產品組合不同而出現顯著差異。本季毛利與毛利率下滑，主要是因為專業服務收入占比提高，而該類收入的毛利率較低。

Our GAAP operating expenses for the third quarter of 2026 were $55.9 million, a year-over-year increase of $25.5 million or 84%. The increase was driven primarily by increases of $12.3 million in personnel costs, $4.1 million in noncash stock-based compensation, $4.6 million in noncash depreciation and amortization, $1.1 million in third-party professional fees, and $1.1 million in marketing expenses. Non-GAAP adjusted operating expenses were $40.5 million, a year-over-year increase of $17 million or 72%.

我們 2026 年第三季的 GAAP 營業費用為 5,590 萬美元，較去年同期增加 2,550 萬美元或 84%。增加主要來自：人員成本增加 1,230 萬美元、非現金股份基礎給付增加 410 萬美元、非現金折舊與攤銷增加 460 萬美元、第三方專業費用增加 110 萬美元，以及行銷費用增加 110 萬美元。非 GAAP 調整後營業費用為 4,050 萬美元，較去年同期增加 1,700 萬美元或 72%。

Net loss for the third quarter was $48.2 million or $0.13 per share, a year-over-year decrease of $91.8 million or $0.28 per share. The decrease was due primarily to $122 million decrease in the amount of noncash, nonoperating charges related to the remeasurement of the company's warrant liability in 2025. As a reminder, all the company's remaining publicly traded warrants were redeemed in November 2025. The adjusted EBITDA loss for the third quarter of 2026 was $37.7 million, a year-over-year increase of $17 million, with the increase primarily due to higher operating expenses to support the company's accelerated product development and go-to-market initiatives.

第三季淨損為 4,820 萬美元或每股 0.13 美元，較去年同期減少 9,180 萬美元或每股 0.28 美元。下降主要是因為 2025 年與公司認股權證負債重新衡量相關的非現金、非營運費用減少了 1.22 億美元。提醒一下，公司所有剩餘的公開交易認股權證已於 2025 年 11 月完成贖回。2026 年第三季調整後 EBITDA 虧損為 3,770 萬美元，較去年同期增加 1,700 萬美元；增加主要是因為為支援公司加速的產品開發與市場推進（go-to-market）計畫而提高了營業費用。

Now I will address the balance sheet and liquidity. As of September 30, 2026, our cash and marketable securities balance totaled $503.4 million, a 40% year-over-year decrease. The decline was primarily due to the use of approximately $253 million of cash to acquire Quantum Circuits in January of 2026. During the third quarter of 2026, the company executed a definitive agreement with the US Department of Commerce, providing D-Wave with access to up to $100 million of US CHIPS and Science Act funding, subject to the achievement of specified milestones. No CHIPS Act funds have been received as of the end of the third quarter.

接下來我將說明資產負債表與流動性。截至 2026 年 9 月 30 日，我們的現金及有價證券餘額合計為 5.034 億美元，較去年同期下降 40%。下降主要是因為我們在 2026 年 1 月收購 Quantum Circuits 使用了約 2.53 億美元現金。在 2026 年第三季，公司與美國商務部簽署具約束力的最終協議，使 D-Wave 得以在達成特定里程碑的前提下，取得最高 1 億美元的《美國 CHIPS 與科學法案》資金。截至第三季末，尚未收到任何 CHIPS 法案資金。

And finally, as we look to Q4 results, a brief reminder of how revenue recognition works for system sales. For example, with the FAU sale, a substantial portion of revenue will be recognized upon delivery, which we expect to occur in the fourth quarter. An additional portion of any system sale revenue is recognized over the installation period, typically several months after delivery. In the case of FAU, that will likely occur in early 2027. There are also small recurring revenue streams related to maintenance and QCaaS access that span several years.

最後，在展望第四季結果時，先簡要提醒系統銷售的收入認列方式。例如就 FAU 的銷售而言，收入中的相當大一部分將在交付時認列，而我們預期交付將發生在第四季。任何系統銷售收入的另一部分，會在安裝期間按期認列，通常是在交付後數個月。以 FAU 為例，這可能會在 2027 年初發生。此外，還有與維護與 QCaaS 存取相關、跨越數年的小額經常性收入來源。

With that, operator, please open the call for questions.

接下來，接線員，請開放提問。

Thank you. (Operator Instructions) John McPeake, Rosenblatt.

謝謝。（接線員指示）John McPeake，Rosenblatt。

Guys, Alan, Greg, welcome to these calls. Nice job there. So questions on on-chip qubit control multiplexing, Alan. That is a bottleneck for the superconducting camp and you guys are working on a solution there. Could you tell us about any progress there or what's happening? And I just had a quick follow-up.

各位，Alan、Greg，歡迎參加這些電話會議。表現得很不錯。我有關於晶片上量子位控制多工（on-chip qubit control multiplexing）方面的問題，Alan。這是超導陣營的一個瓶頸，而你們正在著手解決方案。能否談談目前有任何進展或正在發生什麼？我還有一個簡短的追問。

Sure, John. So as we've discussed in the past, our annealing quantum computers are capable of controlling today over 4,000 qubits with 200 to 300 I/O lines, whereas in the gate-model space, it will typically require 3 to 5 I/O lines per qubit. So we are significantly more efficient on our annealing quantum computers than any gate model system today that requires significantly more I/O lines.

當然可以，John。正如我們過去討論過的，我們的退火式量子電腦目前能以 200 到 300 條 I/O 線控制超過 4,000 個量子位；而在閘模型領域，通常每個量子位需要 3 到 5 條 I/O 線。因此，相較於目前任何需要顯著更多 I/O 線的閘模型系統，我們的退火式量子電腦在效率上高出許多。

The way we're able to accomplish this is by multiplexing control signals over I/O lines and then on-chip being able to decode and route those signals to the appropriate points on the chip. We are carrying this technology over to our gate model systems, specifically, being able to multiplex control signals, decode and route on-chip. As I mentioned a few minutes ago, the first chip design with this new scalable I/O control for gate model systems is out for fabrication, and we expect the chip back this month to be able to begin testing.

我們之所以能做到這點，是透過在 I/O 線上對控制訊號進行多工，並在晶片上解碼並將這些訊號路由到晶片上的適當位置。我們正將這項技術延伸到我們的閘模型系統，具體而言，就是能夠對控制訊號進行多工，並在晶片上解碼與路由。如我幾分鐘前提到的，採用這種新型可擴展 I/O 控制的第一款閘模型系統晶片設計已送交製造，我們預期本月可取回晶片以開始測試。

Okay. And what are your targets there relative to ratio of lines per qubit?

好的。那麼你們在每個量子位所需線路比例方面的目標是什麼？

Yeah, so our goal is to be able to have the same sort of capability that we have on our annealing system, which is to be able to control 100 logical qubits with, a couple hundred I/O lines, 200 to 300. That's our goal. We are working toward that, and this first chip that we are expecting back will allow us to understand if we're going to be able to achieve that goal.

是的，我們的目標是能具備與我們退火系統相同類型的能力，也就是用幾百條 I/O 線（200 到 300 條）來控制 100 個邏輯量子位。這就是我們的目標。我們正朝這個方向努力，而我們預期取回的第一顆晶片將讓我們了解是否能達成該目標。

That's it for me. Thank you.

我這邊就到此為止。謝謝。

Kingsley Crane, Canaccord Genuity.

Kingsley Crane，Canaccord Genuity。

Hi, thanks for taking the question. Congrats on a great quarter. I wanted to just touch again on the extended or perhaps expanded relationship with SkyWater. So I believe Emil Michael had wanted to guarantee that IQ could provide continued access to SkyWater's chips and maybe seek other clauses like firewall or switching assistance.

嗨，謝謝讓我提問。恭喜本季表現亮眼。我想再回到與 SkyWater 延長、或許也可說是擴大的合作關係。所以我相信 Emil Michael 原本希望能保證 IQ 可以持續取得 SkyWater 的晶片，並且可能爭取其他條款，例如防火牆或轉換協助。

And so does your private agreement contractually recreate any of those, or just how would you characterize the partnership at this stage, like extended or expanded, given that you're prudently still evaluating other options for supply chain? Thanks.

那麼你們的私下協議在合約上是否重新建立了其中任何一項？或者你會如何形容目前這個合作夥伴關係的階段——例如是延長或擴大——考量到你們仍審慎評估供應鏈的其他選項？謝謝。

So, first of all, I will say that -- and I've said this before -- the SkyWater team has been totally supportive in working through with us an agreement that, of course, makes sense for both of us, them and us, but gives us confidence that our IP will be protected and not leaked out to other perhaps competitive entities. And also gives us confidence that the highly skilled individuals that are working with us, the highly skilled SkyWater individuals that are working with us on our systems will continue to be a part of the program. So as we've worked through this agreement, we have gained confidence that our working relationship will continue as it had been for the last roughly 10 years.

首先我想說——而且我之前也說過——SkyWater 團隊一直非常支持，與我們一起推進一份協議；當然這份協議必須對雙方都有意義，對他們與我們皆然，但也讓我們有信心，我們的 IP 會受到保護，不會外洩給其他可能具競爭性的實體。同時也讓我們有信心，那些與我們合作的高技能人員——與我們在系統上合作的 SkyWater 高技能人員——將會持續是這個計畫的一部分。因此，隨著我們推進並完成這份協議，我們對於雙方的合作關係能夠延續過去約 10 年的運作方式，已更有信心。

The other thing I will say is that the relationship has now also expanded in a different direction, which is us being a supplier to SkyWater as we have now completed a proof of technology for how to use our annealing quantum computers and hybrid solvers to help optimize their operations. That proof of technology was successful, and now we're moving on to a proof of concept. So we actually now have a really interesting bidirectional relationship where we are supporting one another.

我還想補充的是，這段關係現在也朝另一個方向擴展：我們成為 SkyWater 的供應商，因為我們已完成一項技術驗證，展示如何使用我們的退火量子電腦與混合式求解器來協助最佳化他們的營運。那項技術驗證是成功的，現在我們正進一步進行概念驗證。所以我們現在其實形成了一個非常有意思的雙向關係，彼此相互支援。

Great. Yeah. And looking forward to seeing that proof of concept continue to develop commercially. And then secondly, just as a follow-up, which is, there's been some noise in the broader market post-elections last week. Support for Quantum has been largely bipartisan, but just curious thoughts as we exit those elections and then just any potential changes you think to the existing programs. Thanks.

很好。是的，也期待看到那個概念驗證持續推進並發展成商業化。第二個問題作為追問：上週選後整體市場有一些雜音。對量子領域的支持大致上是跨黨派的，但想請教你們在選舉結束後的看法，以及你認為現有計畫是否可能出現任何變化。謝謝。

So far, I'm not seeing any changes with respect to either the strength of the pipeline or perhaps more important, the sales agreements that are in negotiation right now. So I've seen nothing but continued forward momentum.

到目前為止，我沒有看到任何變化，不論是就管線的強度而言，或更重要的是，就目前正在洽談中的銷售協議而言。所以我看到的只有持續向前的動能。

Okay. Thanks so much.

好的，非常感謝。

Kevin Garrigan, Jefferies.

Kevin Garrigan，Jefferies。

Yeah, hey, Alan and Greg, congrats on all the progress. Hey, Alan, you talked about diversifying your supply chain. Are you seeing any bottlenecks today, whether it be just traditional components or any other quantum components? And what would you look for in terms of potentially starting to think about building inventory for certain components?

是的，嗨，Alan 和 Greg，恭喜各項進展。Alan，你提到要讓供應鏈多元化。你們目前有看到任何瓶頸嗎？不論是傳統零組件或任何其他量子相關零組件？另外，就可能開始為某些零組件建立庫存而言，你們會看哪些指標？

So as we've discussed in the past, in some sense, the largest and most costly single component in our system is the dilution refrigerator. And timelines for that are not as short as we'd like them to be. We don't necessarily see them lengthening, but we'd certainly prefer if we could get dilution refrigerators measured in small numbers a month as opposed to large numbers of months.

如同我們過去討論過的，某種程度上，我們系統中最大、也是成本最高的單一元件是稀釋製冷機（dilution refrigerator）。而它的交期並沒有我們希望的那麼短。我們不一定看到交期變得更長，但我們當然更希望稀釋製冷機的交期能以「每月少量」來衡量，而不是以「好幾個月」來衡量。

The good news, though, is that there are now more suppliers in the market. I mean, typically it's been Bluefors is the leader, and then a couple of others, but now Maybell here in the US is coming on strong, and we've ordered one of their fridges so that we can qualify it as well. So with more suppliers, I'm hopeful that maybe the timelines will actually reduce as opposed to lengthening.

不過好消息是，市場上現在有更多供應商。我的意思是，過去通常是 Bluefors 領先，然後還有幾家，但現在美國這邊的 Maybell 也正在快速崛起，我們已經訂購了他們的一台製冷機，以便我們也能進行資格認證。因此，隨著供應商增加，我希望交期實際上能縮短，而不是拉長。

But other than that, it's more or less business as usual. That having been said, as our pipeline for system sales is growing, as we are making good progress on negotiating new deals, we are starting to buy ahead for some of the longer lead time items.

除此之外，大致上算是照常運作。話雖如此，隨著我們系統銷售的管線在成長、且我們在洽談新交易方面取得良好進展，我們也開始針對一些交期較長的項目提前採購。

Harsh Kumar, BMO.

Harsh Kumar，BMO。

Yeah, hey guys. Alan, I had a quick question and maybe I missed this or didn't understand it properly. I'm hoping you can clarify. You talked about a dual-chip annealing processor that will enable you to sort of a significant move to get to 20K or 100K. Can you just talk about what that is and how it's significant for you to be able to move up and scale?

是的，嗨，各位。Alan，我有個簡短問題，可能是我漏聽了或沒有正確理解。希望你能幫我釐清。你提到一個雙晶片的退火處理器，將使你們能夠大幅推進到 20K 或 100K。你能談談那是什麼，以及它為何對你們提升與擴充規模如此重要嗎？

Sure. So our current generation processor is an Advantage2 system. And up until now, we've grown the number of qubits by putting more qubits on a chip. But we have said that our next generation system, Advantage3, will be a multi-chip solution. Specifically, we will now start scaling by interconnecting multiple QPU processor chips, as opposed to just more qubits per chip.

當然。我們目前這一代的處理器是 Advantage2 系統。到目前為止，我們是透過在單一晶片上放入更多量子位（qubit）來增加量子位數量。但我們也說過，我們下一代系統 Advantage3 將會是多晶片解決方案。

So there are two critical technologies required to be able to do that. One, we need to be able to interconnect those processor chips and preserve the superconducting and quantum properties across them. And then two, we need to ensure that as we do that, we don't have significant growth in the number of I/O lines. So we talked a few minutes ago about the on-chip control technology and using that to keep the number of I/O lines low even as the processor size grows. And that's what that new chip that is in fabrication that we expect back this month is all about. And we're looking at that for both our gate-model processors as well as our annealing processors. We're basically enhancing that control capability to be even more scalable than it has been up until now.

要做到這件事，需要兩項關鍵技術。第一，我們必須能互連這些處理器晶片，並在晶片之間維持超導與量子特性。第二，我們必須確保在這麼做的同時，I/O 線路數量不會大幅成長。所以我們幾分鐘前談到晶片上控制（on-chip control）技術，並利用它在處理器尺寸成長時仍能把 I/O 線路數量維持在較低水準。而這正是那顆正在製造、我們預期本月會拿回來的新晶片的重點。我們同時也在閘模型處理器與退火處理器上評估這項技術。基本上，我們正在強化這項控制能力，使其比迄今為止更具可擴充性。

With respect to the processor interconnect, we have demonstrated already the ability to interconnect what I'll call sample processor chips at a small scale. And we are now in the process of starting fabrication of our first pair of interconnected actual annealing processor chips. Once we do that, we will test to ensure that the interconnect is working, the superconducting bonds are working, the quantum properties are preserved, and we can run that pair of chips as a single integrated processor.

至於處理器互連方面，我們已經在小規模上展示了互連我所稱的樣品處理器晶片的能力。而我們現在正著手開始製造第一對互連的、真正的退火處理器晶片。一旦完成，我們將進行測試，以確保互連正常運作、超導鍵結正常運作、量子特性得以保留，並且我們能把這一對晶片當作單一整合式處理器來運行。

Gary Mobley, StoneX.

Gary Mobley，StoneX。

Good morning, everybody. Thank you for taking my question. I'm not asking for explicit fiscal year '27 revenue guidance, but how do you see the composition of fiscal year '27 revenue changing relative to recent years?

各位早安。謝謝讓我提問。我不是要你們提供 27 會計年度的明確營收指引，但相較於近幾年，你們如何看待 27 會計年度營收組成的變化？

Just thinking in terms of the additional layers, such as pre-sales of 17-qubit gate-model solutions, partnership-driven sales, such as those with CGI, perhaps more Leap quantum computing as a service revenue, more government, which I believe is a new initiative. Any color there would be helpful.

我在想新增的幾個層面，例如 17 量子位元閘模型解決方案的預售、由合作夥伴帶動的銷售（例如與 CGI 的合作）、或許更多 Leap 量子運算即服務（QCaaS）的營收、更多政府端（我相信這是新的倡議）。如果能提供一些補充說明會很有幫助。

Yeah, so let me take a stab at that. First of all, as you know and you said, we don't provide guidance. So I'm not going to provide any numbers or breakdowns, but here's what I will say.

好的，我來嘗試回答一下。首先，如你所知、你也提到，我們不提供指引。所以我不會提供任何數字或拆分，但我可以這麼說。

First of all, we expect that all of system sales, quantum computing as a service, professional service, and sales through partners will grow in 2027 over what we expect we will see by the end of 2026. Now that having been said, because system sales represent significantly larger bookings than either quantum computing as a service or professional services for any single deal, and because when the revenue is recognized on system sales, it typically ends up being recognized over a shorter period of time than is the case with quantum compute as a service, I think you can expect that the mix on bookings will start to be more significant for system sales and the mix on revenue may also be more significant on system sales.

首先，我們預期到 2027 年，系統銷售、量子運算即服務、專業服務，以及透過合作夥伴的銷售，全部都會較我們預期在 2026 年底所看到的水準成長。話雖如此，由於系統銷售在任何單一交易上的訂單金額（bookings）通常明顯大於量子運算即服務或專業服務，而且系統銷售的營收認列通常會在較短期間內完成（相較於量子運算即服務），我認為你可以預期：在訂單組合上，系統銷售的比重會開始變得更顯著；在營收組合上，系統銷售的比重也可能會更顯著。

As far as partners are concerned, we're really just getting started with partners. We signed up a couple of partners that we think could be very helpful to us, and in fact, we're off to a good start. But I think 2027 is a year where we'll start to see that come online as opposed to being a significant component of the business versus the system sales or our QCaaS PS sales.

至於合作夥伴方面，我們其實才剛起步。我們簽下了幾個我們認為能對我們很有幫助的合作夥伴，而且事實上目前進展不錯。但我認為 2027 年會是我們開始看到這部分真正上線、開始發揮作用的一年，而不是相較於系統銷售或我們的 QCaaS/PS 銷售就已成為業務的重要組成。

And then with respect to government, we are seeing a lot more interest from the US government in our annealing systems than we've ever seen. I think that we are making very good progress on a couple of fairly significant application opportunities that I hope to be able to talk about at some point in the relatively near future, and then other opportunities just getting started. So I do think that government is going to be significantly larger component than it has been in the past.

另外就政府而言，我們看到美國政府對我們的退火（annealing）系統的興趣，比以往任何時候都更高。我認為我們在幾個相當重要的應用機會上取得了非常好的進展，希望在相對不久的將來能在某個時間點談到；此外也有其他機會才剛開始。所以我確實認為，政府將會成為比過去大得多的組成部分。

Tyler Ron, Craig-Hallum.

Tyler Ron，Craig-Hallum。

Hi, this is Tyler Ron for Richard Shannon. Thank you for taking my questions. I was wondering if you could talk about NTT's solution, specifically the frequency of which they're running this and whether or not NTT had developed a solution on their own or had used your professional services. And then just as a second question, just to ask it now, any highlights that you would give for what is driving the performance of your solver? Thank you.

你好，我是 Tyler Ron，代 Richard Shannon 提問。謝謝讓我提問。我想請你談談 NTT 的解決方案，特別是他們執行的頻率，以及 NTT 是自行開發解決方案，還是使用了你們的專業服務。另外第二個問題我也先一起問：你們的求解器（solver）表現是由哪些因素驅動的？有沒有什麼重點可以分享？謝謝。

Okay, so relative to NTT Docomo's two production applications, they actually developed both of them on their own. They essentially established a very strong team. They built the first application. We started having some discussions with them. I think maybe we were helpful to them in those discussions, and now they've built a second application, and they're both in production today.

好的，關於 NTT Docomo 的兩個正式上線（production）的應用，他們其實兩個都是自行開發的。他們基本上建立了一個非常強的團隊。他們打造了第一個應用。我們開始和他們進行一些討論。我想在那些討論中我們或許對他們有所幫助，而現在他們又打造了第二個應用，而且兩個目前都已在正式環境中運行。

And honestly, that's great. I mean, we are, first and foremost, in the business of quantum computing, and professional services is a means to an end. We don't really want to be a large PS company. We want to sell systems. We want to sell quantum computing as a service. We use PS to help our customers get to the point where they can get value out of those systems or out of that quantum computing as a service. And so to the extent that partners can help customers do that work, to the extent that customers can do it themselves, that's exactly where we want to be and works well for the model that we're trying to put in place.

坦白說，這很棒。我的意思是，我們首先、最重要的是在做量子運算這門生意，而專業服務只是達成目的的手段。我們並不真的想成為一家大型的專業服務公司。我們想賣系統。我們想賣量子運算即服務。我們使用專業服務來協助客戶走到能從這些系統或量子運算即服務中取得價值的階段。因此，只要合作夥伴能幫客戶完成這些工作，或客戶能自行完成，這正是我們希望達到的位置，也很符合我們正在建立的商業模式。

As far as the two applications, these applications are applications that run relatively infrequently. They are significant in the sense that they can make an important impact to the operations of the company and improve some financial metrics for them, but they don't need to run on a daily or weekly basis. They run more infrequently.

至於這兩個應用，這些應用的執行頻率相對不高。它們的重要性在於能對公司營運產生重要影響，並改善他們的一些財務指標，但不需要每天或每週運行。它們是更低頻率地運行。

With respect to the Stride hybrid solver, most of it was rework of the code to improve the performance of the system. There were some additional capabilities that we incorporated, but a lot of it was rework of the code.

關於 Stride 混合式求解器，大部分是重寫（rework）程式碼以提升系統效能。我們也加入了一些額外功能，但很大一部分是程式碼重構。

Awesome. Thank you.

太好了。謝謝。

Troy Jensen, Cantor Fitzgerald.

Troy Jensen，Cantor Fitzgerald。

Hey, congrats on all the great updates here. Alan, for you, I'd just be curious about Qubits Asia. It feels like in North America here, there's kind of a push for sovereign supply chains. It feels like other countries might do that. But just your thoughts on -- you think that's going to be the case coming back from Asia here?

嘿，恭喜這裡有這麼多很棒的更新。Alan，我想問一下 Qubits Asia。感覺在北美這邊，對於主權供應鏈（sovereign supply chains）有一種推動力。感覺其他國家也可能會這麼做。但想聽聽你的看法——你覺得從亞洲回來後，情況也會是這樣嗎？

Sorry, Troy. I'm not sure I fully understand the question. You broke up a little bit when you were talking about supply chain.

抱歉，Troy。我不確定我是否完全理解你的問題。你在談到供應鏈時訊號有點斷斷續續。

Yeah, just sovereign supply chains. It seems like our administration is really pushing for everything built in the US. Just curious if you're seeing that more globally, especially coming out of this Qubits Asia conference you just did?

對，就是主權供應鏈。看起來我們的政府非常推動所有東西都要在美國製造。我只是好奇你是否看到全球也更朝這方向走，特別是你剛參加完這場 Qubits Asia 會議之後？

So we are seeing it. I mean, the EU, for example, has taken a very strong position that EU quantum funding can go to only EU companies. That's starting to spill over into some of the countries. In Asia, we're not seeing that strong position, but instead what we're seeing is, well, can you at least incorporate some of our companies into your supply chain for some of the components, at least for systems that you deliver here? And we're looking at that.

我們確實有看到。例如歐盟就採取了非常強硬的立場：歐盟的量子資金只能給歐盟公司。這也開始外溢到一些國家。在亞洲，我們沒有看到那麼強硬的立場；但我們看到的是：至少你能不能把我們的一些公司納入你的供應鏈，讓你交付到這裡的系統中某些元件由我們的公司供應？我們正在評估這件事。

We don't -- we do not currently source all our components from the US. And I think while there may be a push from every country to be able to do that, I'm not convinced we'll ever get there. I mean, they're just, certain components where companies in other countries are best at delivering them and other components where local companies may be best at delivering them. So I think for the foreseeable future, the supply chain is going to, by necessity, be more global. And then we just need to manage that. I think we can certainly manage it in the US, we are. I think we can manage it in Asia, we are. I think US -- sorry, Europe is a little bit more challenging. But at least as long as country money doesn't fall into the same track as the EU, we shouldn't have a problem, although as I said a minute ago, it's starting to move in that direction.

我們目前並不是所有元件都從美國採購。且我認為，雖然每個國家可能都會推動能做到這件事，但我不確定我們是否能真正走到那一步。我的意思是，有些元件就是其他國家的公司最擅長供應；也有些元件可能是本地公司最擅長供應。所以我認為在可預見的未來，供應鏈必然會更全球化。然後我們只需要把它管理好。我認為在美國我們當然能管理，我們也正在管理。我認為在亞洲我們也能管理，我們也正在管理。我認為美國——抱歉，歐洲會比較具挑戰。但至少只要各國的資金不走向與歐盟相同的路線，我們應該不會有問題；不過如我剛才所說，它正開始朝那個方向移動。

And so at that point, we start looking at, well, can we get more independent or commercial funds for the systems in Europe? For example, Swiss Quantum Technology is funding the system in Italy, and they're not really a government entity.

因此在那個時間點，我們開始思考：嗯，我們能否為歐洲的系統取得更多獨立或商業資金？例如，Swiss Quantum Technology 正在資助義大利的系統，而且他們並不是真正的政府實體。

Krish Sankar, TD Cowen.

Krish Sankar，TD Cowen。

Yeah, hi, thanks for taking my question. I have two questions, Alan. I'll just ask both of them. How much of the CHIPS $100 million funding can be allocated towards your superconducting gate technology? And the second question is how is the integration progress of the cryo control chip into your annealing systems? Thank you.

是的，嗨，謝謝讓我提問。我有兩個問題，Alan，我就一起問。CHIPS 的 1 億美元資金中，有多少可以分配到你們的超導閘（gate）技術？第二個問題是：低溫控制晶片（cryo control chip）整合到你們的退火系統（annealing systems）的進展如何？謝謝。

So on the second question, as I said a few minutes ago, the first chips that implement the new cryogenic control architecture designed to be more scalable and to be able to start working with both our annealing and our gate model systems are due back this month. So we're just at the point where we should see the first sample chips back and we can start testing them.

關於第二個問題，如同我幾分鐘前提到的，第一批實作新型低溫控制架構的晶片——該架構設計得更具可擴展性，並且能開始同時支援我們的退火與閘模型系統——預計本月會回來。所以我們正處在應該會看到第一批樣品晶片回來、並可開始測試的階段。

As far as the chips funding is concerned, the way we structured it, the work is really on advancing how we design and fabricate and test the chips, the QPUs, and that work can be applied to both, more or less equally. So for the most part, the funding applies to both equally. Although there are some things that are more applicable for the annealing and some things that are more applicable for the gate, but for the most part, what we're doing will apply across both.

至於 CHIPS 資金方面，我們的架構方式是：工作重點其實在推進我們如何設計、製造與測試晶片（QPU），而這些工作幾乎可以同等地應用在兩者上。因此大部分資金會同等適用於兩者。當然也有一些項目更適用於退火、另一些更適用於閘模型，但就大多數而言，我們正在做的事情會同時涵蓋兩者。

Craig Ellis, B. Riley.

Craig Ellis，B. Riley。

Hello, yes, this is Rebecca on for Craig Ellis. It seems like the RPO conversion commentary implies the revenue ramp that's tracking very well. Could you maybe walk us through what's in the pipeline today and what gives you confidence in converting RPO at that pace? And what would need to close -- to get close to or above those levels? Thank you.

你好，是的，我是代 Craig Ellis 提問的 Rebecca。看起來你們對 RPO 轉換的評論暗示營收爬坡（revenue ramp）追蹤得非常好。你們能否帶我們了解一下目前管線（pipeline）裡有哪些內容，以及是什麼讓你們有信心以那樣的速度把 RPO 轉成營收？另外，要達到接近或高於那些水準，需要哪些案子成交（close）？謝謝。

Okay, so we're not going to provide any information on deals in flight and how much -- what the value might be when they might close and so on. As far as the RPO as it exists today, Greg, do you want to maybe take a shot at -- just kind of reiterating what that comprises and how we expect to see it convert to revenue?

好的，我們不會提供任何關於進行中的交易（deals in flight）的資訊，包括金額可能是多少、何時可能成交等等。至於目前的 RPO，Greg，你要不要試著——再重申一下它包含哪些內容，以及我們預期它如何轉換為營收？

Sure, and thanks, Alan. RPO balance has grown significantly due to system sales opportunities as well as the Fortune 100 enterprise deal. So we have good line of sight on when we can deliver those services, whatever they may be. And so we expect 59% to be delivered 12 months after September 30, 2026, and 71% over the 24 months following that September 30, 2026, date.

當然，謝謝你，Alan。RPO 餘額顯著成長，主要來自系統銷售機會以及那筆《財富》100 大企業的交易。因此，我們對於何時能交付那些服務（不論具體是哪些服務）有很清楚的能見度。所以我們預期在 2026 年 9 月 30 日之後的 12 個月內交付 59%，並在該 2026 年 9 月 30 日日期之後的 24 個月內交付 71%。

(Operator Instructions) Quinn Bolton, Needham.

(接線員指示) Quinn Bolton，Needham。

Hey, this is Shadi Mitwalli on for Quinn. Thanks for letting us ask a question. I believe you guys mentioned that you're already in discussions around 17-qubit system sales. Are those engagements with existing and new customers, or are you starting to see growing interest from new customers?

嗨，我是代 Quinn 提問的 Shadi Mitwalli。謝謝讓我們提問。我相信你們提到你們已經在討論 17 量子位（qubit）系統的銷售。這些洽談是與既有客戶和新客戶都有，還是你們開始看到新客戶的興趣在增加？

Discussions with both, existing and new.

既有客戶與新客戶都有在討論。

Got it. And then for my follow-up, you guys highlighted a Fortune 100 customer in professional services. I know you can't say who this customer is, but just any color on how long you've been engaged with them and then maybe what sector and applications they're targeting. Thank you.

了解。那我的追問是：你們強調了一位《財富》100 大客戶的專業服務（professional services）。我知道你們不能說是哪一家，但能否提供一些背景，例如你們與他們合作多久了，以及他們所屬產業、以及鎖定的應用方向？謝謝。

Well, we did talk about this actually in Q1. This was our first kind of two-year, $10 million deal to pursue multiple applications. The first of those is already in production. There are several others that we have been working on with them that are making good progress as well that we believe in the not-too-distant future will move into production. So a lot of the professional services work over the quarter has been in support of that engagement.

嗯，我們其實在第一季就談過這件事。這是我們第一筆為期兩年、1,000 萬美元的交易，用於推進多個應用。其中第一個已經進入正式上線（production）。另外還有幾個我們也一直與他們合作、進展良好，我們相信在不久的將來也會進入正式上線。因此，本季相當多的專業服務工作都是在支援這項合作。

Thank you. I'll now hand the call back to Dr. Alan Baratz for closing remarks.

謝謝。我現在把電話交回給 Alan Baratz 博士作結語。

Okay, so thanks again to all of you for joining us today and for your continued support of D-Wave. We're advancing a clear strategy: expand the commercial use of our annealing technology and execute on our path to fault-tolerant gate model quantum computing. Our focus is on delivering measurable customer value and building a successful and growing business around it. And we look forward to continuing to update you on our progress. Thank you all.

好的，再次感謝各位今天加入我們，也感謝各位持續支持 D-Wave。我們正在推進一個清晰的策略：擴大我們退火技術的商業使用，並沿著我們通往容錯（fault-tolerant）的閘模型量子運算之路執行。我們的重點是交付可衡量的客戶價值，並在此基礎上建立一個成功且持續成長的業務。我們也期待持續向各位更新我們的進展。謝謝大家。

Thank you for your participation in today's conference. This does conclude the program. You may now disconnect. Everyone, have a great day.

感謝各位參與今天的會議。本節目到此結束。您現在可以掛線。各位，祝您有美好的一天。