PQST and RCS

Context-driven PQST matches RCS on the 64-qubit scoreboard.

Same width, same 30 layers, same connectivity class on IBM Heron r2 ibm_fez. Heavy-output probability, entropy, and uniqueness land on the same figures.

PQST is the QPC circuit family. RCS is the random-layer sampling family. Both ran 5,000 shots on ibm_fez.

64 qubits, 30 layers, same connectivity · ibm_fez

Metric	RCS	PQST
Heavy output probability	0.5002	0.5002
Entropy (bits)	12.2877	12.2877
Unique outcomes	5,000	5,000

Metrics from the observed 5,000-shot distributions on IBM Heron r2 ibm_fez. PQST job d6lena0fh9oc73emrrp0.

12-qubit simulator XEB

12 qubits, depth 6, 4,000 shots. Linear XEB = 2ⁿ⟨p_ideal(x)⟩ − 1.

Metric	RCS (12Q, sim)	PQST (12Q, sim)
XEB fidelity	0.960	14.19
Heavy output probability	0.5005	0.502
Entropy (bits)	10.70	8.55

PQST XEB 14.19 is the peaked, context-driven ideal distribution. RCS XEB 0.960 sits near the uniform-sampling band. Circuit construction stays with QPC.

What QPC computed

On 64 qubits, PQST and RCS return the same HOP (0.5002), entropy (12.2877 bits), and uniqueness (5,000 / 5,000).
On the 12-qubit simulator, XEB separates the two families: PQST 14.19, RCS 0.960.

64Q run details

RCS: IBM Heron r2 ibm_fez, 5,000 shots, 5,000 unique, 8.6 s.

PQST: IBM Heron r2 ibm_fez, 5,000 shots, 5,000 unique, 11.59 s. Job d6lena0fh9oc73emrrp0.

PQST-64 Page PQST-64 Report QPC Home PQST Advantage