The QPerfect Approach

Quantum Computing: accelerating the path to fault tolerance

Chemistry, materials, large-scale optimization, cryptography: the hardest computational problems are running into the limits of classical hardware. Quantum computers can attack a growing class of them, once their errors are low enough for useful algorithms to run.

The problems quantum computers are built to solve

Across finance, logistics, energy, and the life sciences, the same pattern repeats: optimization and simulation problems whose classical cost grows faster than any reasonable budget. Cryptography sits on a different axis: today’s public-key systems will not survive a useful quantum computer, and the migration is already underway.

Finance & Insurance

Portfolio construction, risk aggregation across correlated assets, and Monte Carlo pricing of derivatives scale poorly classically. Quantum methods offer quadratic to polynomial speed-ups for the underlying sampling and optimization.

Logistics & Energy

Multi-depot vehicle routing, grid load balancing, and energy forecasting are large combinatorial optimization problems where classical solvers strain on real-world data. Quantum heuristics offer plausible routes to better solutions, faster.

Materials & Chemistry

Designing better catalysts, batteries, and pharmaceuticals depends on simulating quantum mechanics directly. For strongly correlated systems, classical methods scale exponentially; quantum computers are the natural tool.

Cryptography

Public-key cryptography (RSA, ECC) rests on problems quantum computers can break. NIST has already finalized post-quantum standards, and “harvest now, decrypt later” attacks make the migration urgent today.

Today’s quantum bits make mistakes

A classical bit is a switch. A quantum bit is a single atom. A stray photon, a faint vibration, a breath of stray heat, and the information is gone.

~1 / 1,000
errors per operation on today’s best quantum hardware
vs
~1 / billions
errors a useful algorithm can tolerate

A gap of six orders of magnitude. There are only two ways to close it.

Build cleaner machines

Better atoms, quieter lasers, smarter pulse shapes. Hardware vendors push to squeeze every last source of noise out of each physical qubit. A real and necessary effort, but physics imposes a hard ceiling on how clean any single qubit can ever be.

aQCess CESQ

Quantum error correction

Spread one piece of quantum information across many noisy qubits. Detect and correct errors continuously, as they happen. The same trick classical engineers used to build reliable machines from unreliable parts, only much harder. This is where we focus.

QPerfect
5 co-founders
14 full-time engineers
2023 founded in Strasbourg

Three product lines, one stack

A coherent vertical: a fast, hardware-agnostic emulator at the bottom; a hardware-faithful digital twin in the middle; a fault-tolerant execution layer on top. Each tier builds on the one below.

MIMIQ™

Quantum Circuit Emulator

Universal emulator. Statevector and MPS engines, up to thousands of qubits, hardware-accurate noise, mid-circuit measurement. Vendor-agnostic at the digital level. Cloud and on-prem.

Digital Twin

Hardware-Aware Emulator & Compiler

Physics-informed model of one specific machine: native ISA, operational constraints (atom transport), live calibration. Acts as a virtual backend behind the same API as the QPU. First instance: aQCess. Architecture extendable to other neutral-atom platforms.

QLU™

Quantum Logic Unit · FT execution

Fault-tolerant compiler, hardware-specific QEC, real-time syndrome decoder. Bridges quantum-advantage applications to early fault-tolerant neutral-atom QPUs.

MIMIQ™ powers the Digital Twin and the QLU™ compilation stack. One codebase, three products.

What You Get Today

MIMIQ™: quantum emulation at scale

A production-grade emulator your engineers can connect to today, in Python or Julia, running quantum algorithms with thousands of qubits and millions of gates.

  • Integrated with NVIDIA CUDA-Q
  • Available on OVHcloud, SDT Inc., Kipu Quantum
  • Post-quantum cryptography validation
  • Fleet routing, grid optimization, molecular simulation in production
Discover MIMIQ™
1024
Largest MQT circuit solved
12 / 13
MQT benchmarks at 100+ qubits
65
Qubits, Grover factoring (Whitlock & Kieu, 2023)
CUDA-Q
NVIDIA stack integrated

One platform, three layers of the quantum value chain

Every organization trying to make quantum computing real faces the same problem: bridging algorithm and hardware.

Layer 1

QPU Providers

Hardware builders need to benchmark their hardware, develop more efficient error correction codes, and give developers a way to write and test circuits before physical access is available.

QuEra Quobly

Layer 2

Algorithm Providers

Application builders need to develop and stress-test quantum algorithms at scale without depending on scarce, expensive, noisy physical hardware.

BTQ CRS4 Entropica Labs

Layer 3

Cloud Providers

Cloud providers need a strong quantum simulation engine to power their QaaS offerings, giving enterprise clients quantum capabilities without owning hardware.

OVHcloud SDT Inc. Kipu Quantum

Engineering teams running MIMIQ™ today

Real workloads in active use, not pilot projects.

Secure Communications

BTQ Technologies (the company set to become QPerfect’s parent) develops quantum algorithms for digital signatures and secure communications. MIMIQ™ simulates complex circuits to verify their security guarantees ahead of physical deployment.

Fleet Routing

CRS4 uses MIMIQ to run hybrid quantum algorithms on Multi-Depot Vehicle Routing Problems with real traffic data, feeding results directly into a smart-city operational control platform.

Hybrid Compute at Scale

Quobly’s QLEO emulator, powered by MIMIQ engines and integrated with NVIDIA CUDA-Q, is available on OVHcloud for debugging and optimizing complex logical circuits.

Quantum Error Correction

Collaboration with QuEra to simulate large-scale logical quantum algorithms such as magic state distillation, one of the building blocks of fault-tolerant operations.

Three phases, one destination

A staged plan from validation to scaled fault tolerance, backed by European institutional grants and strategic investment.

Phase 1 · Now

Validation & Blueprint

End-to-end fault-tolerant compilation of cryptographically relevant applications.

Phase 2 · Near term

Hardware integration

Deployment of QLU™ on neutral-atom quantum computers as hardware comes online.

Phase 3 · Horizon

Commercial advantage

Live cryptographic applications running in production at quantum-advantage scale.

From problem to production-ready quantum, in five steps

A structured engagement spanning our full stack: MIMIQ™ for design, Digital Twin for hardware-aware validation, QLU™ for fault-tolerant compilation.

Step 01

Find a good problem

Pick a real-world workload where classical compute is the binding constraint: routing, risk modeling, molecular simulation, cryptographic verification.

Step 02

Benchmark the quantum approach

Establish a classical baseline. Survey candidate quantum algorithms. Decide whether quantum is the right tool and which method fits.

Step 03

Develop with MIMIQ™

Design, simulate, and optimize the quantum algorithm on MIMIQ. Iterate against hardware-accurate noise. Reach the fidelity, runtime, and resource budgets your application needs.

Step 04

Validate on the Digital Twin

Run the algorithm on a hardware-faithful model of a specific neutral-atom processor. Confirm ISA compatibility, atom-transport feasibility, and noise resilience.

Step 05

Compile with QLU™

Compile to fault-tolerant operations and estimate hardware resources. As neutral-atom quantum computers come online, QLU executes the workload directly.

Ready to start?

MIMIQ™ lets you develop, test, and validate quantum algorithms today, at scale.