• 

  Just an AI image generated for this challenge

Inspiration

The YQuantum 2025 BlueQubit Challenge inspired us to: learn what peaked circuits are, what are their specialities, work with BlueQubit's SDK and last but not least deal with optimization of circuits. Our goal was to democratize quantum advantage by making circuits faster, smaller, and more error-resilient—unlocking potential in fields like cryptography, chemistry, and machine learning.

What it does

Team Deteqt's project delivers a hybrid optimization framework that:

• Reads and extracts important information like total qubit number and different gate number from the .qasm file
• With the optimization techniques, it reduces the gate counts and circuit depth
• Also deliver the important graphs which can be used to optimize the framework more and understand the circuit

How we built it

Our team at first started using the introductory jupyter notebook shared by the instructors which uses IBM's workspace, we implemented the base logic into BlueQubit's SDK. BlueQubit's cpu and mps.cpu was able to run until third problem, so we decided to swap to IBM's workspace.

Toolstack:

• Qiskit
• BlueQubit
• Graph-based Algorithms

Methodology:

• Optimization Level 1: Gate/depth reduction via unitary synthesis and pattern matching
• Optimization Level 2: Layout transpilation for IBM hardware using SWAP insertion heuristics

Challenges we ran into

First of all, the first challenge we ran into was error noise. In which we realized that was happening because of aggressive gate reduction which was also increasing our error rate. Second, after the third problem we were facing a memory problem which qiskit_aer simulator couldn't handle, we also then swapped to IBM's workspace.

Accomplishments that we're proud of

We achieved of 35% at average gate reduction. So we are definitely proud of that, We were working on an SDK which was unknown to us, we are feeling great that we've learnt it now. We also did a 3D design of the problem circuits.

What we learned

We learned a new SDK to work with. We tried new ways to optimize our solution. Also, that hardware definitely matters in these type of optimizations.

About our project

Please don't forget to visit out github repo link down below to check our solution!

Built With

• bluequbit
• jupyter-notebook
• python
• qiskit