Introduction

The Quantum Innovation Challenge consists of two phases. During Phase 1, you will form your team, and you will submit your proposal. After the evaluation of the proposals, the best five proposals will enter a Phase 2, where the teams will have the opportunity to access the Gefion AI Supercomputer. The access to Gefion will be shared among the team members.

The Danish Centre for AI Innovation (DCAI) runs and operates Gefion, a large-scale NVIDIA DGX SuperPOD. It comprises 191 NVIDIA DGX H100 systems for a total of 1,528 NVIDIA H100 Tensor Core GPUs (more information at https://dcai.dk/). Project submission requirements The projects must be submitted via a GitHub pull request where your code repository is licensed under the MIT (http://opensource.org/licenses/MIT) or Apache Version 2.0 license (http://www.apache.org/licenses/LICENSE-2.0). If your submitted repository has no license affiliated with it, then your submission will be regarded as invalid and not taken into evaluation.

Eligibility

To be eligible for participating in the Quantum Innovation Challenge, the following requirements must be met:

  1. Teams must consist of an academic member and start-up members.
  2. The academic member must be employed or affiliated with a university, hospital, or other non-profit research institution and will receive the voucher grant for Gefion access during Phase 2. The institution will serve as administering institution for the grant.
  3. Each team member must have an active role. Each team is responsible to define the role of each team member.
  4. The academic member must send a hosting letter signed by the head of department of the administering institution to quantum_challenge@mqs.dk.
  5. Teams may submit only one proposal for this challenge.

Intellectual Property (IP)

  1. The projects are “pre-competitive”.
  2. Start-ups participating in the competition and the industry partners involved in organizing the initiative cannot receive patentable IP funded by the financial sponsors.
  3. Any foreground knowledge generated under the project must not be protected by any form of intellectual property rights, including patents, and cannot be kept confidential or treated as trade secrets.
  4. All foreground knowledge must be made publicly available as soon as reasonably practicable. The start-up and the academic member involved in the Challenge, the industrial partners involved in organizing the initiative and third parties will have the same right to use the foreground knowledge without restrictions.
  5. The academic institutions, start-ups and industrial partners are entitled to conduct research based on the foreground knowledge. Any results, findings and/or data generated outside the scope of the project stemming from such research may be subject to intellectual property protection, including patenting, if applicable. However, the project teams and industrial partners involved in the Challenge initiative are prohibited from seeking any intellectual property protection on any results, findings or data generated outside the scope of the project but derived from or based on the foreground knowledge generated during the project before the foreground knowledge has been published.
  6. Any subsequent IP based on the projects must be based on additional funding and research will not be funded by the financial sponsors of the Challenge initiative.
  7. Any knowledge generated by the projects will be made publicly available.

Language

Documentation regarding the challenge must be submitted in English.

Application process

When all proposals have been assessed, teams will be notified about whether they have been selected among the five teams. The notification e-mail will be sent from quantum_challenge@mqs.dk. PLEASE NOTE: The scientific committee of Quantum Innovation Challenge does not provide feedback in case a project submission is declined as one of the top five projects. But there will be efforts to share a feedback and rating document to all participating teams for maintaining transparency and valuable feedback for future developments of the submitted project.

Assessment criteria

1. Problem Understanding & Approach (0-5 pts)

  • Comparison to state of the art: Does the team clearly define the problem and put it into context?
  • Technical Feasibility: Is the approach reasonable given current or near turn quantum computational/AI hardware capacities?
  • Innovativeness: Does the solution bring a novel perspective or improvement compared to existing approaches? Has the participant demonstrated (potential) advantage over classical methods (e.g. speedup, accuracy, resource efficiency)?
  • Scalability: Can the approach be extended or scaled up? For larger PKPD datasets can this solution adapted for another use case such as drug resistance prediction?

2. Computational Resource Estimation (0-5 pts)

  • Hardware Requirements: Did the team estimate the needed computational resources (GPUs, CPUs)?

3. Technical Assessment (0-5 pts)

  • Choice of Algorithms: Are the selected algorithms appropriate for the task?
  • Justification of Methods: Is there a solid reasoning behind using specific architectures or techniques?
  • Benchmark with existing solutions: Are existing solutions considered and benchmarked?

4. Structure of the Work Plan (0-5 pts)

  • Step-by-Step Breakdown: Does the participant outline a clear roadmap how to approach the challenge?
  • Risk Assessment & Mitigation: Are potential failure points identified with alternative solutions?

5. Team skills 0-5 (pts)

  • Skillset: Does the team cover the necessary experience in quantum/AI techniques?

6. Proof of Concept / Prototype (non-mandatory; 0-5pts)

  • Prototype Implementation: Has the team built a minimal working proof of concept?
  • Initial Results: Are preliminary results provided?
  • Demonstration of Key Concepts: Does the proof of concept validate core ideas effectively?