Kvantify has successfully scaled its FAST Variational Quantum Eigensolver (FAST-VQE) algorithm to 50 qubits on IQM’s Emerald quantum computer, demonstrating a significant advance in practical quantum chemistry simulations. Utilizing the Qrunch software suite, the company performed calculations exceeding the capabilities of classical computation, operating at IQM’s facilities with the commercially available Emerald processor. This achievement—building on earlier 16- and 20-qubit studies—highlights FAST-VQE’s scalability, maintaining constant circuit counts while exploring chemical systems inaccessible to conventional methods, and demonstrating faster convergence compared to random gate selection, even with hardware noise.
Scaling Quantum Chemistry with 50 Qubits
Kvantify recently scaled its FAST-VQE quantum chemistry algorithm to 50 qubits on IQM’s Emerald processor, a significant step beyond prior “toy problem” simulations. This achievement allows modeling of chemically relevant systems—like the butyronitrile dissociation reaction—with active spaces exceeding the capabilities of classical CASCI methods. The 50-qubit scale creates a Hilbert space too large for direct classical simulation, demonstrating a practical advantage for quantum computation in tackling complex chemical problems currently inaccessible to traditional methods.
A key innovation is Kvantify’s FAST-VQE algorithm, designed for scalability by maintaining a constant circuit count, unlike other VQE variants. On IQM Emerald, FAST-VQE consistently converged faster than random gate selection, even accounting for hardware noise. Results show the algorithm narrows the computational space effectively during initial operator selection, a critical step for speeding up calculations. This demonstrates today’s quantum hardware can capture structural information and deliver results beyond what randomness provides.
The butyronitrile dissociation study revealed a shift in bottlenecks as scale increases. While quantum execution remains crucial, classical optimization of operator parameters became the limiting factor. A “greedy” optimization strategy—adjusting one parameter at a time—yielded a 30 kcal/mol energy improvement compared to full-parameter optimization, highlighting the need for algorithmic advancements on the classical side. This signals that future progress in quantum chemistry will require co-development of both quantum hardware and optimized classical workflows.
FAST-VQE Performance and Scalability
Kvantify recently demonstrated the FAST-VQE algorithm’s scalability by running it on IQM’s 50-qubit “Emerald” processor. This represents a leap beyond proof-of-concept, enabling calculations for chemically relevant systems exceeding the capabilities of classical computation. Specifically, the team tackled the dissociation of butyronitrile, a problem where the required Hilbert space is too large for traditional methods. This achievement highlights a critical shift – quantum hardware is now capable of capturing structural information that classical computers struggle to simulate, even with inherent noise.
A key innovation driving this scalability is FAST-VQE’s constant circuit count, unlike some VQE variations that see exponential growth with system size. Kvantify’s approach combines adaptive operator selection directly on the quantum hardware with an approximate, chemistry-optimized classical simulator for energy estimation. Benchmarking against random gate selection revealed that the quantum hardware consistently converged faster, demonstrating a clear advantage beyond chance. This suggests that even with present-day noise, quantum devices can identify patterns inaccessible to purely classical approaches.
As hardware scales, bottlenecks are shifting from quantum execution to classical optimization. In the butyronitrile study, optimizing all operator parameters became computationally expensive, limiting progress. Kvantify overcame this by employing a “greedy” optimization strategy—adjusting parameters one at a time—achieving 120 iterations daily, compared to only 30 with full-parameter optimization, and delivering a ~30 kcal/mol energy improvement. This demonstrates that future advancements will require a focus on refining classical algorithms to effectively harness growing quantum resources.
Results from Butyronitrile and Future Implications
Recent work by Kvantify, utilizing IQM’s 50-qubit Emerald processor, demonstrates a significant leap in quantum chemistry simulations. Specifically, the dissociation of butyronitrile – a molecule too complex for classical computation at this scale – was successfully modeled using the FAST-VQE algorithm. This achievement isn’t simply about qubit count; it signifies reaching a point where quantum hardware can deliver insights beyond the reach of purely classical methods, opening doors to modeling increasingly complex chemical systems.
A key finding revolves around shifting computational bottlenecks. While quantum execution remains a challenge, the study revealed that classical optimization of parameters is now the limiting factor as hardware scales. A “greedy” optimization strategy—adjusting one parameter at a time—improved energy calculations by approximately 30 kcal/mol compared to full-parameter optimization, highlighting the need to focus on optimizing the classical components of hybrid algorithms.
Importantly, Kvantify’s FAST-VQE consistently outperformed random gate selection, demonstrating the value of structured quantum computation even with noise present. This proves that today’s quantum devices aren’t just performing calculations; they’re capturing structure within the chemical problem. Scaling to 50 qubits with IQM Emerald is therefore a crucial step, proving that quantum chemistry is moving beyond proof-of-concept and towards practically useful simulations.
