IBM has announced Nighthawk r3, its next-generation quantum processor, built to run 10,000 quantum operations accurately across 520 qubits. The goal is to tackle larger and more complex problems while keeping the speed and quality of the current Nighthawk r2.

Psst!Add AlexTech.ai asPreferred Source on Google

The key figure for industry watchers is scale. In 2027 IBM plans to link two Nighthawk r3 chips into one 1,040-qubit system, past the 1,000-qubit mark. To get there, the company must show that individual chips can be joined without losing reliability.

official image of IBM Quantum Nighthawk r3

Image: IBM — https://newsroom.ibm.com/2026-11-11-ibm-quantum-nighthawk-demonstrates-new-levels-of-speed-and-quality,-with-plans-to-scale-beyond-1,000-qubits-in-2027

Why speed matters as much as accuracy

A quantum computer returns its answers as samples: each run of a circuit yields a single result, so useful answers require millions or billions of repetitions. The most demanding applications need circuits with thousands of operations on around a hundred qubits or more. According to Jay Gambetta, director of IBM Research, usefulness depends on delivering results that are both accurate and fast.

Nighthawk r2, introduced in September, delivers up to 100,000 samples per second and supports accurate computations with more than 7,500 operations on 120 qubits. Its fast qubit-reset technique shortens the gap between executions. Clients and partners including E.ON, Algorithmiq and Q-CTRL have run 34 billion circuits on it, with speedups of up to 19 times over earlier IBM processors. According to these sources, E.ON reported a 15-fold speedup in quantum machine learning for energy grid optimization.

This builds on earlier work. The verifiable quantum advantage study with the University of Chicago already showed that quantum results can be checked; the new generation aims to make that kind of verification repeatable at larger scale.

How two chips become one

Nighthawk r3 keeps the square-lattice layout and fast operations of its predecessor, but adds ten chip-to-chip connections, called l-couplers, designed to link processors into larger systems. IBM has already tested the principle: earlier this year it connected two Nighthawk r1 chips with four l-couplers, reaching 99.3% quantum state transfer fidelity against a 99.9% target. It also joined two modular cryogenic cells, reaching operating temperature in under five days and cooling below 15 millikelvin, as already reported for the quantum refrigerators.

IBM has also revised its plan. As Ars Technica reports, the number of linked systems has dropped from nine to two, while the total qubit count for next year stays the same: fewer chips to join, each one larger.

Software stays in the middle

IBM argues the challenge is not only physical. Developers must combine error correction, noise mitigation and classical post-processing, and that integration will need more classical resources. To handle it, the IBM Quantum Platform now unites high-performance classical computing and quantum computing in one managed environment, from a laptop to a supercomputer.

According to the company, a workflow developed with RIKEN researchers ran roughly five times faster while performing a hundred times more computations. Cleveland Clinic and RIKEN automatically processed 608 million samples to model biologically relevant structures, 33 times faster than without workflow automation.

The next concrete step is 2027. Until then, the route toward fault-tolerant systems runs through making several processors work together, not just adding qubits to a single chip.