China announced the Tianyan-504 superconducting quantum computer with a 504-qubit Xiaohong chip. This is Xiaohong 1.
Summarize this article with:
A 2% CZ error rate roughly means that 1 in 50 two-qubit gates will have an error so you could possibly get by running a 50 qubit circuit of depth 1 (lol), but there is absolutely no reason (beyond marketing and maybe proving something about yield and ability to route wires) to have 500 physical qubits—you can’t even run the shortest possible circuit on them. There’s a caveat that, if you could do error correction, the above argument would no longer be quite right. However, if your error rate is above some threshold (and 2% is, by a lot), error correction won’t correct anything and will just make things worse. "CTQG, which spearheaded the project, claimed that the Xiaohong chip rivals international competitors like IBM in key performance metrics, including qubit lifetime (how long a qubit can hold its quantum state) and readout fidelity (accuracy in extracting information from qubits). Although no link was provided in the statement to research that would verify these claims. These benchmarks are critical for reliable and scalable quantum computing." Exciting announcement but definitely looking forward to hearing more about the benchmarks. 504 qubits They definitely arent error corrected though, are they They definitely arent error corrected though, are they It's pretty standard for companies to just report the number of physical qubits, if you zoom in on the blurry image it looks like they claim 5 logical qubits I don't think anyone in the world can currently claim they have ANY number of EC logical qubits that can perform universal QC. Is this like a VPS? How is it used? Or is this just like "hey look what we can do as China" flex Create your account and connect with a world of communities.