--- title: IBM Brisbane Quantum Processor url: https://www.emergentmind.com/topics/ibm-brisbane-quantum-processor type: topic --- # IBM Brisbane Quantum Processor Searching arXiv for IBM Brisbane and closely related IBM Quantum processor papers to ground the article in current literature. IBM Brisbane Quantum Processor is IBM’s **127-qubit superconducting quantum processor** identified in the literature as `ibm_brisbane`, and it appears in arXiv publications as a large-scale cloud-accessible platform for executing NISQ-era protocols involving entanglement dynamics, unitary-channel discrimination, and broader hardware-aware circuit benchmarking [2311.15973], [2505.17731], [2508.05331]. Within the cited works, Brisbane is characterized not by a single canonical benchmark suite but by its use as an experimental substrate for structured multi-qubit circuits, topology-constrained native-gate compilation, and repeated-shot studies of noise, variability, and measurement behavior. The available literature presents Brisbane simultaneously as a processor for fundamental quantum-information experiments and as a testbed for assessing the interaction between connectivity, entanglement overhead, calibration stability, and circuit depth on present-day superconducting hardware [2311.15973], [2505.17731]. ## 1. Processor identity and architectural role In the cited literature, IBM Brisbane is presented as a **127-qubit superconducting quantum processor** and as part of IBM’s cloud-accessible hardware stack for circuit execution and hardware-aware benchmarking [2311.15973], [2508.05331]. One paper explicitly refers to `ibm_brisbane` as a processor that supports **linear chains of connected qubits** suitable for embedding nearest-neighbor circuits, while another describes Brisbane as a **third-generation IBM quantum chip** used for performance estimates, calibration comparisons, fidelity calculations, and error-correction trend analysis [2311.15973], [2508.05331]. The architectural picture that emerges is functional rather than exhaustive. The papers do not provide a full device specification sheet, but they do identify operationally significant properties. Brisbane’s connectivity is sufficiently structured that five-qubit line subgraphs can be selected directly for experiments, and the device’s native compilation behavior matters enough that manual qubit placement, fixed logical-to-physical mappings, and topology-aware construction are repeatedly discussed [2311.15973], [2505.17731]. This suggests that Brisbane is best understood as a processor whose practical behavior is strongly shaped by routing constraints and native entangling-gate orientation, rather than as an abstract 127-qubit register with uniform all-to-all access. A plausible implication is that Brisbane’s importance in the literature lies less in nominal qubit count alone than in the combination of scale, cloud accessibility, and sufficient controllability to support simultaneous sub-experiments, hardware-native transpilation studies, and structured many-qubit discrimination circuits. ## 2. Native-gate and topology constraints A recurring theme across Brisbane-specific experiments is that implementation choices are dominated by native-gate availability and directed connectivity. In the multiple-shot unitary-discrimination study, the authors emphasize that **CNOT is not a native gate** on Brisbane and therefore construct **ECR-based circuits** to better match the hardware [2505.17731]. The same work states that **ECR gates are oriented**, so qubits cannot be swapped arbitrarily without affecting compilation strategy and depth [2505.17731]. This constraint appears concretely in circuit families that compare CNOT-based and ECR-based realizations. For larger circuits, the authors report that direct logical-to-physical mapping is only feasible for a few qubits; otherwise one must either manually rearrange the circuit for each size or introduce swap gates, which increase depth [2505.17731]. The distinction is not merely formal. Their data indicate that hardware-aware ECR-based designs can outperform naïvely transpiled CNOT-heavy constructions, especially in larger-width settings [2505.17731]. The open-system entanglement experiment on `ibm_brisbane` provides an independent example of topology-constrained synthesis. Because the processor connectivity did not allow a direct CNOT from \(q_1\) to \(q_3\), the circuit used the chain \[ q_1 \to q_2,\qquad q_2 \to q_3,\qquad q_3 \to q_2 \] to realize the required effective interaction [2311.15973]. The native gate set there included the echoed cross-resonance gate (ECR), \(RZ\), \(X\), identity, and \(\sqrt{X}\), with \[ \sqrt{X}=\frac{1}{2}\begin{pmatrix}1+i & 1-i \\ 1-i & 1+i\end{pmatrix}. \] The same paper writes the single-qubit preparation rotation as \[ \sqrt{X}\,RZ(\lambda)\,\sqrt{X} = \begin{pmatrix} \sin(\lambda/2) & \cos(\lambda/2) \\ \cos(\lambda/2) & -\sin(\lambda/2) \end{pmatrix}, \] again illustrating how Brisbane experiments are often expressed directly in the hardware’s native or near-native gate vocabulary [2311.15973]. Taken together, these reports indicate that Brisbane’s effective computational model is inseparable from its directed entangling primitive and layout-aware compilation regime. This suggests that algorithmic performance on Brisbane is often determined as much by embedding strategy as by the ideal logical circuit. ## 3. Experimental uses on `ibm_brisbane` Brisbane has been used for at least two distinct classes of experiments in the cited literature: open-system entanglement dynamics and multiple-shot unitary-channel discrimination. A third paper treats Brisbane as the reference processor for broader circuit benchmarking involving Jaynes–Cummings-inspired and longitudinal Ising-model circuits [2508.05331]. In "Quantum simulation of entanglement dynamics in a quantum processor" [2311.15973], the device hosted a **five-qubit protocol** in which two qubits represent the main system, two represent the environment, and one is an ancilla used to estimate concurrence. The processor allowed **three independent 5-qubit experiments simultaneously** on disjoint chains such as \[ \{q_0,q_1,q_2,q_3,q_4\},\qquad \{q_6,q_7,q_8,q_9,q_{10}\},\qquad \{q_{27},q_{28},q_{29},q_{30},q_{31}\}, \] which the authors used to probe both parallelism and qubit-subset variability [2311.15973]. The initial two-qubit entangled state was prepared as \[ |\psi\rangle=\sin(\lambda/2)\,|00\rangle+\cos(\lambda/2)\,|11\rangle, \] with amplitudes corresponding to \[ \alpha=\frac{1}{\sqrt{2}},\qquad \frac{1}{\sqrt{3}},\qquad \frac{1}{\sqrt{5}}, \] run simultaneously on the three chosen subsets [2311.15973]. In "Experimental study of multiple-shot unitary channels discrimination using the IBM Q computers" [2505.17731], all experiments were run on **IBM Quantum processor Brisbane**. The paper studied two discrimination tasks: first, discrimination between \(\Phi_{\mathrm{id}}\) and \(\Phi_{RZ(\phi)}\) with no intermediate processing; second, discrimination between \[ U=\sqrt{X}\,RZ\!\left(-\frac{\pi}{2N}\right)\sqrt{X},\qquad V=\sqrt{X}\,RZ\!\left(\frac{\pi}{2N}\right)\sqrt{X}, \] with hardware-friendly processing inserted between uses [2505.17731]. The circuits spanned **pure parallel**, **pure sequential**, and **hybrid rectangular** architectures of width \(w\) and depth \(d\), with \(N=wd\), and the study explicitly reports Brisbane performance on **6-qubit systems**, **11-qubit systems**, and larger hybrid settings [2505.17731]. In "Quantum Circuit Benchmarking on IBM Brisbane: Performance Insights from Superconducting Qubit Models" [2508.05331], Brisbane is the platform for comparing a **Jaynes–Cummings-inspired circuit** and a **Longitudinal Ising-model circuit**. The latter is explicitly described as a **22-qubit circuit** with layers of Hadamard gates followed by phase \(P\) and \(R_z\) rotations and then adjacent-qubit CNOT gates [2508.05331]. That work frames Brisbane as a processor for evaluating frequency behavior, fidelity decay, and error-correction demand under increasingly large structured workloads. ## 4. Entanglement dynamics and open-system simulation The most detailed physics application of `ibm_brisbane` in the cited literature is the simulation of entanglement transfer in an amplitude-damping setting [2311.15973]. In the representative chain \(\{q_0,q_1,q_2,q_3,q_4\}\), the mapping was: system qubits \(q_1,q_3\), environment qubits \(q_0,q_4\), and ancilla \(q_2\) [2311.15973]. Each system qubit was coupled to an independent reservoir qubit under the master equation \[ \dot{\rho}_S=\gamma\sum_{j=\{1,3\}} \left[ \sigma_{-,j}\rho_S\sigma_{+,j} -\frac{1}{2}\{\sigma_{+,j}\sigma_{-,j},\rho_S\} \right], \] with the associated map \[ |0_{q_{1(3)}}0_{q_{0(4)}}\rangle \rightarrow |0_{q_{1(3)}}0_{q_{0(4)}}\rangle, \] \[ |1_{q_{1(3)}}0_{q_{0(4)}}\rangle \rightarrow \eta\,|1_{q_{1(3)}}0_{q_{0(4)}}\rangle+ \zeta\,|0_{q_{1(3)}}1_{q_{0(4)}}\rangle, \] where \[ \eta=e^{-\gamma t/2},\qquad \zeta=\sqrt{1-e^{-\gamma t}}. \] After tracing out the reservoirs, the reduced system state takes the X form \[ \rho_S(t)= \begin{pmatrix} s^2+c^2\eta^4 & 0 & 0 & sc\,\eta^2 \\ 0 & c^2\eta^2\zeta^2 & 0 & 0 \\ 0 & 0 & c^2\eta^2\zeta^2 & 0 \\ sc\,\eta^2 & 0 & 0 & c^2\zeta^4 \end{pmatrix}, \] with \[ c=\cos(\lambda/2),\qquad s=\sin(\lambda/2). \] Because of this X structure, the concurrence was written as \[ \mathcal{C}_S=\max\left[0,-2\eta^2(c^2\zeta^2-sc)\right]. \] The ancilla-based protocol converts concurrence estimation into a probability measurement. Using the Bell-like state \[ |\Phi\rangle=\sqrt{\frac{1}{2}}\left(|00\rangle+|11\rangle\right), \] the paper defines \[ P_{\Phi}=\langle \Phi|\rho_S|\Phi\rangle = \frac{1}{2}\left[1-2\eta^2(c^2\zeta^2-cs)\right], \] so that \[ \mathcal{C}_S=\max\{0,2P_\Phi-1\}. \] Operationally, the relevant quantity is extracted from the three-qubit outcome \(|010\rangle\) on \((q_1,q_2,q_3)\), with \[ \mathcal{P}_{010}=\frac{P_\Phi}{2}, \qquad \mathcal{C}_S=\max\left\{0,4\mathcal{P}_{010}-1\right\}. \] For the environment qubits, the same witness-based procedure was used after swap operations through the chain [2311.15973]. The experimental findings were qualitative but physically specific. For \(\alpha=1/\sqrt{2}\), system entanglement decays asymptotically while environment entanglement grows from \(t=0\). For \(\alpha=1/\sqrt{3}\), the system exhibits **sudden death of entanglement** at finite time \(t_d\), whereas the environment exhibits **sudden birth of entanglement** at an earlier time \(t_b