QInnovision World Challenge 2025
Pioneering Quantum Solutions for Industry Transformation
Hosted by
QInnovision
Posted
Our AMA with Nvidia's @esperanzacuencagomez is Starting in One Hour! 🎙️
Please comment your questions below 👇
Order by:
Posted by keeswagemans •
Hi Esperanza, Thank you for taking the time to answer some questions for us :) What format of interaction with Nvidia is planned for future advancements? Best regards, Kees
Posted by esperanzacuencagomez •
Thanks a lot for your question, Kees. I am not sure I understand your question. You mean in the context of this event or in any other context? Thanks!
Posted by keeswagemans •
I am sorry the ambiguity, but I mean within this challenge!
Posted by esperanzacuencagomez •
Thanks for the clarification! Just start a discussion on the discussion channel in this platform, tag me and I'll be happy to help :-)
Posted by bernalde •
Hi Esperanza, thank you for this chance! How will you guys evaluate solutions to this challenge that do not directly use the QAOA implementation by CudaQ but other tools provided by NVidia?
We ended up submitting two different approaches, one with Cuda QAOA and another one (the one we are more excited about) using a new TensorSolver that can use CUDA in NVidia GPUs.
Cheers, SECQUOIA
Posted by esperanzacuencagomez •
Two solutions using NVIDIA technology and tools are great! We encourage using CUDA-Q and we are happy to see other NVIDIA tools too. Also, please note that for this challenge CUDA-Q is our recommendation. Hope this helps!
Posted by kraji •
Hi @esperanzacuencagomez, I’d love to hear your thoughts on the future of metrics and scoring systems for evaluating quantum solutions. How do you see these evolving to ensure fair benchmarking and meaningful comparisons as the field matures?
Posted by esperanzacuencagomez •
That's a great question and one to which I really don't have an answer. I personally believe that those scoring systems and metrics are hard to define at this stage beyond the clasical metrics for qubit and gate fidelity, coherence, and so on. Anything beyond that it's very hard for me to answer at this point.
Posted by kirill_shiianov •
Hello Esperanza, and thank you for providing this QnA for us! :)
My first question would be about the use-case and data:
For the second stage of the competition, could the competitors (who get there, of course!) get any realistic data to work with? Would you have an indication to a 'preferred open-source' provider?
Thank you kindly!
Kindest regards,
Kirill
Posted by esperanzacuencagomez •
Sure, happy to Kirill! Thanks for your question. We are not providing data for the challenge and we don't have any indications or preferences for open-source providers, so we are happy to leave that to your best judgement - which I am sure it will be good.
Posted by kirill_shiianov •
Thank you very much, Esperanza!
Posted by keeswagemans • (edited)
Hi Esperanza, Would it be possible to connect with someone involved in the development CUDA-Q? We couldn't enjoy the capabilities of CUDA-Q in it's full potential because of the available documentation. We see some innovative solutions, but we can't advance with these, because of the same reason. Meeting with a developer would help us greatly.
Posted by esperanzacuencagomez •
Can you share more details of the challenges you are facing? I am happy to review those with our engineers if needed, though a meeting with them can be hard to schedule. However, if you share more details, maybe we can discuss those through an off-line channel. I am also happy to take that meeting myself if that's helpful. Thanks!
Posted by keeswagemans •
We are facing issues with the ambiguous definition of some internal functions of the cudaq.vqe. What would be your preferred way of reaching out to you?
Posted by esperanzacuencagomez • (edited)
You mean about the cudaq.vqe Phyton API? Where do you see ambiguity and how could that ambiguity be solved?
Maybe this VQE example can be helpful: https://nvidia.github.io/cuda-quantum/latest/applications/python/vqe.html, as well as the VQE documentation in our recently released CUDA-QX libraries: https://nvidia.github.io/cudaqx/components/solvers/introduction.html#variational-quantum-eigensolver-vqe
Please feel free to reach out to me through a discussion thread in this platform.
Thanks a lot for your questions!
Posted by keeswagemans • (edited)
I have added two files with two errors we received while running cudaq.vqe. One with optimizer GradientDescent and the other one with optimizer COBYLA. Is there a possibility we could schedule a call to show our code to someone of NVIDIA? Before the 1st of February if that fits your schedule. COBYLAError.txt GradientDescentError.txt
Posted by esperanzacuencagomez •
Thanks a lot for sharing these errors. We are looking into those. Unfortunately, we cannot schedule a call to review the code before the 1st of February, but would you guys be open to share the code in any other way? Maybe a private GitHub would work?
Many thanks
Posted by kirill_shiianov •
My second question would be the following:
For further exploration of scaling of the developed solutions, could the second-phase contestants possibly get access to any cloud computing resources by Nvidia?
CudaQ has shown very high potential on our local machines, so it might be curious to explore how high can a well-optimised hybrid routine go on a scalable backend.
Thank you very much, and looking forward to a chance to collaborate with Nvidia!
Kindest regards,
Kirill
Posted by esperanzacuencagomez •
Thanks a lot for the follow up question, Kirill. We are not providing cloud computing resources for this challenge, but I encourage you to explore cloud access by other means (Google Colab for example)
Posted by RMTboy •
Hi Esperanza, thanks for hosting this AMA! I’m curious—how do you see competitions, benchmarking, and open-source initiatives shaping the quantum computing ecosystem, similar to how they’ve driven innovation in AI?
Posted by esperanzacuencagomez •
Thanks a lot for the question. I believe that all of those are great initiatives to drive the quantum computing ecosystem forward. At NVIDIA, we are following an open-source strategy with CUDA-Q, our open-source platform for accelerated quantum supercomputing. CUDA-Q is an open-source quantum development platform orchestrating the hardware and software needed to run useful, large-scale quantum computing applications. The platform’s hybrid programming model allows computation on GPU, CPU, and QPU resources in tandem from within a single quantum program. You can learn more here: https://developer.nvidia.com/cuda-q and here: https://nvidia.github.io/cuda-quantum/latest/index.html
Posted by RMTboy •
I have seen the release of your recent survey on the use of AI for quantum computing, would you be open to some form of competition in this area one day potentially ? This could be a way to bridge the two communities.
Posted by esperanzacuencagomez •
It's something that we haven't really think about but it is an interesting idea for sure. Thanks a lot for sharing that!