AI Sovereign Compute Infrastructure Program (SCIP) is sponsored by Innovation, Science and Economic Development Canada (ISED). Canada is investing in a new, state-of-the-art, Canadian-owned and located AI supercomputing system to support researchers and a cross-section of industry. This program aims to exponentially increase the compute power currently available in Canada.

Program guide: Artificial Intelligence Sovereign Compute Infrastructure Program (SCIP) Program guide: Artificial Intelligence Sovereign Compute Infrastructure Program (SCIP) a.

Sovereignty requirements Section 1: Applicant information Section 2: List of partners Section 3: Applicant and partners profile Section 4: Application summary Section 5: Technical requirements Section 6: Sovereignty requirements Section 7: Other application requirements Section 8: Source of funds Section 9: Estimated costs Section 10: List of attachments Section 11: Authorization and certification a.

Accessing the form and deadline Technical execution and priorities assessment Detailed description of program layers This Program Guide is intended to assist Applicants in the completion of the Application Form for the Artificial Intelligence Sovereign Compute Infrastructure Program (SCIP).

It provides information on the program, the Application Form, the process for submitting the application, as well as the criteria for assessing the application. It is strongly recommended that Applicants read this Guide before filling out the application form to ensure a thorough understanding of both the program objectives and Applicant eligibility.

The application form must be completed and submitted via email to aiscip-picsia@ised-isde. gc. ca by 1:00 PM Eastern Time on Monday, June 1, 2026 , with the required information as outlined in this Guide.

During the review process, Innovation, Science and Economic Development Canada (ISED) program officials may engage with A pplicants directly as required.

Under the Financial Administration Act and the Treasury Board's Policy on Transfer Payments and associated Directive , federal departments may only commit and expend public funds where Parliament has provided the necessary spending authority through appropriations, and no payment or financial commitment may lawfully be charged against an appropriation in excess of or without that authority.

Accordingly, the launch of SCIP and any subsequent contribution agreements are contingent upon the department receiving the required appropriations and authorities from Parliament and the Treasury Board.

The issuance of this call for applications does not constitute a legal or financial obligation on the part of the Government of Canada to enter into a contribution agreement, make any payment, or fund any proposal unless and until such authorities are obtained and a formal contribution agreement is executed in accordance with applicable legislation and policies governing transfer payments.

Applicants are encouraged to also review the SCIP webpage for more information on the program. If there are any additional questions, please contact the program at: aiscip-picsia@ised-isde. gc.

ca . The objective of SCIP is to vastly increase the total AI-relevant compute power currently available to Canadian researchers and industrial research and development (R&D) by building state-of-the-art AI supercomputing infrastructure.

This program aims to: Meet the compute needs of Canadian researchers, giving them the tools to undertake transformative AI projects; Support access for innovative Canadian businesses; Enhance Canadian sovereignty and resiliency; Allow for the integration of Canadian technologies; and Accelerate the build-out of national public AI compute capacity to ensure Canada remains globally competitive and retains top AI talent.

SCIP is guided by a set of core priorities that underpin the Government of Canada's strategic investment in building a large-scale, sovereign AI compute capability. Proponents are expected to demonstrate how their proposals contribute to these priorities, which will advance Canada's capacity to support world-leading research and innovation.

Increasing compute capacity: Expanding the availability of high-performance, AI-optimized compute infrastructure in Canada to meet growing demand from the research and innovation ecosystem. Speed to delivery: Enabling the rapid deployment of meaningful compute capacity to ensure timely access for users and to maintain Canada's competitiveness in a fast-evolving global landscape.

Future scalability: Designing infrastructure that can grow and adapt over time, with consideration for modular expansion, long-term sustainability, and evolving technology needs.

Sovereignty and governance: Establishing a Canadian-located, Canadian-governed system that ensures data residency, operational control, and decision-making authority and agency remain in Canada, and which can be adapted and scaled over time to serve the evolving needs of Canadians.

Economic impact: Supporting broader economic benefits for Canada, including enabling industrial R&D, strengthening domestic supply chains, fostering collaboration, and developing highly qualified talent. Please see the Application assessment in this Guide for information on the assessment process and assessment criteria.

The delivery of SCIP will be structured into two distinct layers: an Infrastructure Build Layer and a National Service Layer. The two distinct layers reflect the need to both establish a sovereign AI supercomputing system and ensure its effective integration into Canada's existing broader digital research infrastructure (DRI) ecosystem.

The National Service Layer will focus on user support and ensuring the new system integrates smoothly with existing public infrastructure access and allocation systems, and that applicable accessibility and cybersecurity standards are met.

To ensure national coordination and seamless access for users across Canada, the successful recipient of the Infrastructure Build Layer will be required to work closely with the National Service Layer provider to fully integrate the new compute infrastructure into the existing national DRI ecosystem through the National Service Layer.

This collaboration will be expected to enable interoperable access, alignment with established governance, security and data management frameworks, and compatibility with existing user support, allocation, and identity systems, thereby ensuring that the infrastructure operates as a coherent component of Canada's broader research and innovation platform rather than as a standalone asset.

Under SCIP, the National Service Layer Provider will facilitate the integration of the SCIP system into the national DRI ecosystem, and will provide user support, data services, training, and research consultation services to the Public Research and Industrial R&D users of the SCIP system.

The Build Recipient will be expected to collaborate with the National Service Layer provider in areas such as AI compute services, research software & tools, security, compliance & trust, knowledge sharing and governance. In addition to collaborating in these key areas, the Infrastructure Build Layer will include systems administration, hardware operations, and data centre operations.

The figure below shows the services scoped to the Build Layer in blue, the services scoped to the Service Layer in green, and the overlap where shared responsibility and collaboration for specific services is expected. A Venn diagram comparing the Build Layer and Service Layer. The left circle, labeled Build Layer, lists systems administration, hardware operations and datacentre operations.

The right circle, labeled Service Layer, lists research consulting, training and skills development, user support and data services. The overlapping section labelled Collaboration lists AI compute services; research software and tools; security, compliance and trust; knowledge sharing and governance, showing the areas of collaboration between the Build Layer and the Service Layer.

This call for applications is only for the Infrastructure Build Layer. See the Detailed description of program layers in this Program Guide for more details on the distribution of services between layers. d.

Funding available Footnote 1 The program may provide up to approximately $890 million to the Infrastructure Build Layer to support the design, construction, and ongoing operation of the compute system, over seven fiscal years, beginning in fiscal year 2026–27 as outlined in this call for applications. a.

Sovereignty requirements Sovereign AI compute infrastructure in the context of SCIP is defined as a Canadian-located, Canadian-governed system that ensures data residency, operational control, and decision-making authority and agency remain in Canada, and which can be adapted and scaled over time to serve the evolving needs of Canadians, ensuring: Data sovereignty and protection: Canadian data, including sensitive data, has appropriate protections and can be leveraged to build AI solutions that directly address Canadian needs and priorities.

Data is managed prioritizing data residency in Canada. Infrastructure control: Core compute and storage infrastructure must be owned or contractually controlled by Canadian entities, with safeguards that limit instances in which a foreign party can unilaterally restrict use or access.

Governance and access: Decision-making on system components, and who can use the infrastructure and for what purposes must rest with Canadian institutions or organizations, not foreign vendors or governments. Controlling Canadian entities must attract and retain sufficient deep technical expertise and audit rights to give agency to their decision-making accountability.

Buy Canadian and supply chain resilience: While some components (i.e. hardware, software etc.) may be foreign-made, supply contracts should, when possible, maximize Canadian integration, installation, as well as support, and limit lock-in risks. To uphold SCIP's sovereignty requirements, only organizations that meet the criteria described below will be considered.

Applicants are encouraged to review these guidelines thoroughly to ensure their organization aligns with the eligibility requirements: Not-for-Profit Organizations incorporated in Canada; Post Secondary Institutions incorporated in Canada; Consortia (Led by a Not-for-Profit or Post-Secondary Institution): Multi-partner collaborations that may include academic institutions, research organizations, and industry partners, provided the Lead Applicant is a not-for-profit or post-secondary organization incorporated in Canada.

The Lead Applicant is the organization designated to represent the consortium. If the application is successful, the Lead Applicant would be responsible for signing the Contribution Agreement with ISED. As a reminder, this call-out is only for the Infrastructure Build Layer of SCIP.

Please refer to the Detailed description of program layers in this Guide for more information on the distribution of services for the two layers.

Eligible activities Footnote 2 for SCIP (including both the Service and Build Layers) are those activities related to enhancing public compute infrastructure, data and talent development, and include: Activities to build and enhance compute infrastructure, including maintaining, operating and upgrading compute capacity and exploring innovative approaches; Activities to provide Canadian researchers and industry with access to compute resources; Activities to test novel hardware and software solutions to improve systems performance; Activities to develop sovereign data storage solutions; Activities to utilize, protect, and leverage Canadian data; Activities to advance the safe, ethical, and responsible use of AI technologies; Activities to support research data management Activities to build capacity and attract, retain, train and develop highly qualified and industry-ready personnel; Collaborative activities, including participation of Canadian researchers and innovators in international projects; Day-to-day operations as they relate to the fulfilment of the terms of the contribution agreements; Activities to support the creation, growth and retention of companies; Activities to lease compute infrastructure may be considered but Applicants must demonstrate how it will meet the sovereignty requirements.

Eligible costs are those that relate to and are deemed eligible by the Minister to advance the program objectives and eligible activities. Eligible costs for activities in support of SCIP may include: Expenditures related to the building, leasing, purchasing, maintenance and operations of compute infrastructure.

Leased compute infrastructure may be considered but Applicants must demonstrate how it will meet the sovereignty requirements; Expenditures related to data curation and usage; Salaries, stipends, and benefits for employees' portion of time spent on the project; Operations of facilities; Professional and technical services; Costs associated with researcher, industry and innovation engagement; Expenditures related to networking and communication: e.g., seminars and workshops; networking meetings; conferences; communications; public outreach; Expenditures related to talent development training and capacity building; Reasonable operational expenditures.

In most instances operational expenditures will be limited up to 15%, with the possibility of up to 20%: e.g., administrative salaries; benefits and management fees; goods and services of contractual personnel; consulting and/or technical services; translation, office supplies/services; insurance; staff travel; accommodation; rent and utilities; audit.

As these expenditures depend on the nature and scope of the Recipient's activities, they will be specified in the contribution agreements; Travel should be in line with the Modern Travel Practices of the National Joint Council Travel Directive (NJC). Eligible hospitality costs will follow the standards set out in the current Treasury Board Directive on Travel, Hospitality, Conference and Event Expenditures (THCEE).

Applicants are encouraged to specify any expenditures related to the purchase and/or leasing of land and buildings necessary for their project. However, expenditures related to land and/or buildings are contingent upon the department receiving the required appropriations and authorities from Parliament and the Treasury Board.

Any costs incurred by the Applicant prior to being formally confirmed as a successful recipient are undertaken at the Applicant's own risk. The eligibility of retroactive costs is not guaranteed and, if considered, will only be addressed during the negotiation of the Contribution Agreement. The following information provides detailed explanations and the level of detail expected for each section of the application form.

Section 1: Applicant information 1. 1 Organization information Only applications from eligible organizations will be considered.

Applicants must select their organization type from the following list of eligible organization types: Not-for-Profit Organizations incorporated in Canada; Post-Secondary Institutions incorporated in Canada; Consortium (Led by a Not-for-Profit or Post-Secondary Institution): Multi-partner collaborations that may include, for instance academic institutions, research organizations, and industry partners, provided the Lead Applicant is a Not-for-Profit Organization incorporated in Canada or a Post-Secondary Institution incorporated in Canada.

1. 2 Applicant information Applicants must provide details about their legal and operating names, location, mailing address, official language, and date of incorporation. Applicants must attach proof of Not-for-Profit status or proof that their organization is a Post-Secondary Institution .

For Not-for-Profit organizations, this proof may include incorporation documents (articles of incorporation or certificate of incorporation), CRA Notification of Registration letter, annual returns filed with Corporations Canada or respective provincial registry, or by-law governing documents. For Post-Secondary Institutions, this proof may take the form of a copy of university charter, provincial statute, or ministerial consent.

Applicants must confirm whether they are acting as the Lead Applicant for the purpose of the Application. Applicants must also specify the number of employees (part-time and/or full-time) who will be dedicated to the SCIP project at the Lead Applicant's organization and the estimated annual hiring of personnel dedicated to the SCIP project at the Lead Applicant's organization. 1.

3 Authorized application contact An authorized application contact must also be identified, and their contact information must be provided. The authorized application contact is the person who is authorized to represent the organization and/or consortia and who will be the main point of contact regarding the application.

If the person is not a signing officer of the organization, additional details about the person's role within the application and confirmation of their right to represent the organization should be provided. Section 2: List of partners Applicants must list all organizations who are partners in the application.

Partners are organizations (including academic institutions, research organizations, civil society, industry partners, other levels of government) who, along with the Lead Applicant, will be responsible for achieving the objectives of the project.

For each partner, Applicants must provide the organization's name, business number, or Quebec Enterprise Number (NEQ), location and the number of employees (part-time and/or full-time) who will be dedicated to the SCIP project and the estimated annual hiring of personnel dedicated to the SCIP project.

If there are more than three partners, Applicants must attach information about additional partners (preferably a Word document or PDF), including their name, business number/NEQ, and location. Section 3: Applicant and partners profile 3. 1 Applicant and partners profile Applicants must provide an overview of the qualifications of the key staff who will be involved in the SCIP project.

This may include relevant prior experience, experience in delivering high-performance compute or AI systems, education, publications, and professional designations. Applicants must also describe their financial management capabilities and those of any partners who may be involved.

This may include the organization's current financial management systems and controls, ability to manage and track a program budget, processes to ensure partners direct program funding to eligible costs, and the ability to steward ISED funds. Applicants must attach the Lead Applicant's Financial Statements from the past three years.

In the case where the Lead Applicant is unable to provide Financial Statements they are required to contact the program to discuss alternative documentation. Applicants must provide details about any current or pending funding agreements the Lead Applicant has with any federal department. Applicants must describe their staffing plan for the build, deployment, and ongoing operations.

This includes providing details about key roles, responsibilities for each role, and describing who will be accountable for development, testing, and deployment activities. Applicants must attach a detailed staffing plan of the personnel who will operate, optimize and maintain the infrastructure. Applications must clearly describe the role and contributions of every partner involved in the project.

This includes a detailed explanation of the project's governance structure, outlining decision-making processes, accountability mechanisms, and how partners will coordinate across all phases of the project.

Applications must also explain how the governance structure will ensure that decision-making authority will rest with Canadian institutions, including sufficient deep technology audit rights (i.e. NDA access to proprietary code) and domestic operational expertise.

This ensures that domestic experts have the ability to review possible changes to firmware or software to ensure the Applicants and their partners have the agency to detect risk, and make and enforce decisions.

In addition, Applicants must outline their organization's current and ongoing interactions within the broader AI and supercomputing ecosystem which should include a description of the organization's role in ecosystem level governance.

Applications need to demonstrate how these ecosystem connections and governance structures will support effective partnerships, knowledge exchange, and long-term sustainability of the supercomputing infrastructure. Section 4: Application summary Applications must have an executive summary that highlights the key aspects of the project.

Section 5: Technical requirements The following requirements outline key considerations for hardware, site facility, energy, cooling, security, cybersecurity, scalability, and sustainability that must be addressed in full to ensure the infrastructure can support large-scale AI research and innovation.

ISED recognizes that there have recently been significant price increases for state-of-the-art GPUs due to supply chain issues, and there is the potential for future price volatility due to ongoing supply chain constraints.

Applicants are encouraged to consider how the proposed architecture can maximize the number of advanced GPUs, integrated with the necessary interconnect, storage, cooling, and power infrastructure, to deliver high-performance AI compute while remaining consistent with the program's delivery timeline requirements.

Facilities must be able to receive hardware and commence service operations in an expedient manner ensuring that Canadian researchers have the tools to stay competitive. Applicants must demonstrate a viable plan to ensure hardware delivery and significant service offerings targeting 18 months following entering into a Contribution Agreement.

Applicants must describe the proposed system architecture and expected performance of the infrastructure. This includes providing details on the overall system architecture, including compute nodes (GPUs and CPUs), interconnect, memory, storage, and networking, and how these components are integrated to support AI and HPC workloads. Applicants must include anticipated performance metrics and benchmarking.

Applicants must also explain how the architecture is modular and designed to scale, including maximum supported GPU configurations, expansion pathways, and anticipated performance impacts as the system reaches full capacity and in the case of future expansion in the event of future investment.

If the proposed infrastructure hardware and software are a close match to an existing, well-documented operational system, attach copies of its documentation, and name it, and describe all major improvements or deviations .

If the proposed infrastructure is not a close match to any existing well-documented operational system, explain how and why such novelty will meet SCIP’s priority for the rapid build-out of high-density AI compute to meet the immediate and growing needs of Canadian researchers and innovators.

Provide a description of a representative sampling (at least three) of workloads/use cases relevant to Canadian researchers and industry R&D that this equipment will be appropriate for, including widely available benchmark suites, configurations and performance targets, applicability in different research areas, and, to delineate the scope of research anticipated, a description of workloads that would not be appropriate for this equipment at this scale.

Applications must indicate the location of the site(s) intended to house the infrastructure. Applicants must clearly describe the ownership and legal status of the site, including both the land and any buildings that will be used for project delivery. This includes a full explanation of who owns the property, the nature of any leases or shared use agreements, and any third party interests.

If ownership is held through a corporation, partnership, trust, or other structure, Applicants must identify the ultimate controller of the property. Where a foreign state, entity, or individual has any direct or indirect ownership interest in the ultimate controller, this must be disclosed in detail to support transparency and due diligence requirements.

Applicants must also outline a realistic and detailed plan to achieve significant service delivery within approximately 18 months of entering a contribution agreement and target of when full operational capacity will be achieved.

Include a timeline of major milestones, such as site preparation, construction, installation, commissioning including system acceptance test and benchmarking plans, refresh strategies and phased hardware (including GPUs, CPUs, storage, and memory components) acquisition approaches.

Indicate which permits and approvals are required, the status of each (e.g., pending submission, submitted, approved), and identify any foreseeable supply chain or procurement challenges. Outline strategies, including proof of pre-approved agreements with suppliers, for ensuring timely access to equipment, materials, labour, and professional services as well as compliance with environmental, sustainability, and regulatory obligations.

Applicants must attach technical design specification documents. This may include architectural or engineering drawings, environmental and power system specifications, manufacturers' technical documentation for installed equipment such as CPUs/GPUs, HVAC equipment specifications, rack and containment system specifications .

Applicants must attach a detailed plan (i.e. Work breakdown structure, procurement plan including pre-approved supplier agreements, Gantt chart, resource plan, etc.). Applicants must describe the proposed operational environment and software stack that will support large-scale AI and HPC workloads.

This includes describing the middleware and operational software stack that will be used to deploy, operate and manage the AI compute infrastructure at scale. Applicants need to include details of the tools and systems that will support day-to-day operations such as orchestration, scheduling, monitoring, automation, configuration management, and performance management.

In addition, Applicants must describe the user-facing software stack(s) that will be made available to academic researchers and industrial R&D users. Applicants need to explain how this software stack(s) will be able to support multiple environments tailored to different use cases and security requirements and able to meet the needs of researchers and industry R&D.

At all levels of the software stack, Applicants are encouraged to consider open-source software solutions where appropriate and in keeping with SCIP’s objectives. Applicants must provide a comprehensive risk assessment that supports the operational plan including power and water availability at full operational levels.

Identify key risks that could affect the project's schedule, cost, compliance, or operational readiness, along with mitigation strategies to reduce the likelihood or impact of each risk. Outline contingency plans that explain how the project will remain on track in the event of delays, permitting challenges, supply chain disruptions, system downtime, or other unforeseen issues.

A strong submission will demonstrate that the Applicant has proactively assessed uncertainties and established feasible, well-supported approaches to maintaining project momentum. Applicants must demonstrate that their facility can meet both immediate and long-term energy requirements.

Indicate the site's current available power (in MW) and the power that will be available for significant service targeting 18 months following Contribution Agreement execution. Indicate the power (in MW) that will be available for full deployment as well as potential future power (in MW) that may be available post full deployment. Demonstrate how the design will support growth of power capacity by full deployment.

Responses should explain planned upgrades or agreements that will enable the site to scale as required. Describe their current and future power provisioning and management arrangements, including electricity sources, load management strategies, redundancy measures, and any sustainability or efficiency features.

This should include details on how the design and planned infrastructure will support growth to full deployment outlined within the scope of this call and future expansion. Include information about grid connection plans, utility coordination, planned infrastructure upgrades, and any sustainability or energy-efficiency features incorporated into power design.

To verify the energy supply, Applicants must attach evidence of committed or planned power arrangements. Acceptable documents may include power purchase agreements, utility connection or capacity confirmation letters, electricity supply contracts, or feasibility studies from the utility provider. Documentation should clearly support the timelines and capacity levels described in the application.

Applicants must describe the facility's current and planned cooling capacity, explaining how the system will support the computing loads for the site. Provide an estimate of the annual volume of water (in millions of litres) drawn and discharged. Explain how the cooling system can scale to meet future demand as supercomputing requirements evolve.

This includes planned upgrades, modular components, or pathways for increasing capacity over time. Outline the cooling technologies being used, how they are optimized for energy efficiency, and any design features that reduce overall power usage (e.g., heat recovery systems, advanced airflow management, liquid cooling, or free air cooling where applicable).

Applicants must describe the facility's physical security measures, including systems designed to prevent, detect, and respond to security incidents. Include fire prevention and suppression systems, controlled personnel access (such as secure entry points, authentication measures, or restricted zones), and onsite surveillance or monitoring technologies.

Outline disaster recovery safeguards in place to maintain continuity of operations during emergencies, such as backup power, redundant systems, incident response procedures, or secure data handling protocols. 5.

8 System design and cybersecurity Applicants must describe the measures in place to protect the cybersecurity of the infrastructure, including how the system will support multiple, purpose-built user environments with varying security requirements.

Outline how environments for open academic research, controlled industry use, and sensitive data (such as health-related information) will be segregated and governed, with clearly defined access controls, authentication processes, and cybersecurity measures aligned with data sensitivity and applicable Canadian and international standards, regulatory requirements, and best practices.

Include details on existing and planned data storage infrastructure, such as secure storage tiers, encryption practices, and data management policies to demonstrate that the system can protect data across all operational contexts. 5. 9 Scalability and emerging technologies Applicants need to clearly explain how their system and site design enable scalability over time.

Explain how the physical site, facility design (e.g., power, cooling, and networking infrastructure), and system architecture could accommodate future hardware upgrades, technology refreshes, integration of emerging technologies (e.g. quantum computing) and capacity expansion to maintain relevance and competitiveness of the system.

Applicants must describe the energy efficiency of the facility, including the current or planned average Power Usage Effectiveness (PUE). Outline the energy procurement processes and policies, demonstrating how the facility will minimize its environmental footprint, for example, through renewable energy sourcing, load-shifting strategies, or other low-carbon approaches.

Describe any broader sustainability initiatives, such as water-conservation measures, waste-heat reuse, circular-economy practices for hardware lifecycle management, or participation in recognized environmental certification programs. Together, these elements should demonstrate a comprehensive and forward-looking approach to sustainable facility operations. Section 6: Sovereignty requirements 6.

1 Data sovereignty and protection Canadian data, including sensitive data, must have appropriate protections, while being leveraged to build AI solutions that directly address Canadian researcher needs and priorities. Outline how the application will prioritize data residency in Canada and ensure data management practices comply with Canadian privacy and security laws without imposing burdens that could impede researcher success.

Explain how cross-border data access or legal issues will be identified, mitigated and managed in a way that avoids imposing onerous security measures on research that does not involve specifically sensitive data, while ensuring that carefully prescribed protections are applied when sensitive data is required. 6.

2 Infrastructure control and access Core compute and storage infrastructure must be owned or contractually controlled by Canadian entities, with safeguards to ensure a foreign party cannot unilaterally restrict use or access. Decision-making on who can use the infrastructure and for what purposes must rest with Canadian institutions or organizations, not foreign vendors or governments.

Controlling Canadian entities must attract and retain sufficient deep technical expertise and audit rights to give agency to their decision-making accountability. Outline who will own and control the core compute and storage infrastructure across the full lifecycle of the facility and how this controlling body will have sufficient deep technical skills to achieve agency.

Describe how the Applicant will ensure that decisions related to access to infrastructure will remain under the authority and control of Canadian institutions and organizations. Specify any policies, measures, or agreements to ensure foreign entities cannot restrict use or access. 6.

3 Buy Canadian and supply chain resilience To ensure a secure, resilient supply chain that prioritizes the involvement of Canadian vendors and technology providers in all stages of the project while making room for emerging technologies, Applicants must: Include plans to integrate Canadian technologies, products, solutions, and services across hardware, software, and operational layers wherever possible and feasible.

Demonstration of existing relationships with key suppliers/partners will be seen as an asset. Outline how the system will be designed and built to enable future integration of innovative technologies, such as emerging AI accelerators, quantum components, or advanced interconnects. This will ensure the system remains adaptable and capable of incorporating next-generation hardware and software solutions.

Integrate approaches to foster innovation through partnerships with Canadian firms, including start-ups. Section 7: Other application requirements 7. 1 Knowledge of Research Ecosystem Applicants must describe how their organization demonstrates a comprehensive understanding of the needs and operating realities of Canadian researchers, as well as the diverse AI workloads required across both academic and industry users.

This may include knowledge of sector-specific compute demands and the performance, scalability, and reliability requirements of a broad range of AI use cases. 7.

2 Access and Cost Recovery Models SCIP is intended to anchor a national AI-focused computing system to meet the needs of researchers, in addition to industrial research and development, with an aim to accelerate discovery, strengthen collaboration between academia and industry, and support innovative Canadian businesses conducting advanced research and development.

Applicants must demonstrate: A plan to provide a portion of compute access for industry. This could include consideration around paid proprietary industry access, including a transparent cost-recovery model and reinvestment plan.

Applications should outline how revenue generated from commercial access might be reinvested to sustain and expand the system, for example, funding system operations and maintenance, and supporting future hardware refreshes to maximize long-term national benefit. A plan to enable collaborative research between industry and academia. Applications need to explain how the project strengthens the Canadian