HPE merges Cray supercomputers with quantum systems: a new stage of hybrid computing
Hewlett Packard Enterprise (HPE) has made a strategic move by expanding its coalition of partners to integrate classic Cray supercomputers with quantum computing systems. The alliance includes key players in the quantum industry such as Intel, IQM, Qblox, Quantinuum, QuEra Computing, Quantum Machines, Rigetti, and Riverlane. The goal is to create full-fledged hybrid architectures that combine the power of traditional HPC systems with promising quantum processors.
Project Essence: From Theory to Practical Testbeds
HPE is not just announcing collaboration — the company intends to physically integrate the HPE Cray platform with quantum processors, qubit control systems, and error correction solutions. Partners have already begun building specialized testbeds. These platforms will serve as proving grounds for jointly developing hybrid algorithms, testing software compatibility, and evaluating the real-world performance of such systems. This is not about laboratory experiments but about creating prototypes capable of solving specific computational tasks.
Why This Matters for the Market
Quantum computing cannot yet replace classical computing — it is only effective for a narrow range of tasks, such as optimization and cryptography. The hybrid approach, where a quantum accelerator operates under the control of a powerful classical supercomputer, is considered the most realistic path to practical application. HPE, owning the Cray platform (one of the most powerful in the world), holds a unique position to build such bridges. The involvement of Intel, Rigetti, and Quantinuum ensures access to different types of qubits — from superconducting to ion traps — increasing development flexibility.
My view: This move by HPE is not just a marketing alliance but a pragmatic bet on the next 3–5 years. Integrating quantum processors with Cray supercomputers could provide the necessary infrastructure for the first real quantum-classical applications. However, it is critically important how quickly the partners can solve the error correction problem — without this, hybrid systems will remain demonstration testbeds rather than working tools.