You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!
Will TSM dominate fabricating quantum chips, like it dominates standard chips? If not, what companies do the readers feel will be the dominate players in this area and will TSM compete? Also will new types of memory be enabled by quantum structures?
Cheap answer - you could argue they already are since transistors have been experiencing quantum effects for a few nodes now (and only TSMC has the most experience with yields at these class of nodes): https://semiengineering.com/quantum-effects-at-7-5nm/
There are many different platforms to build quantum technologies and they all have very different requirements. It isn’t clear which platform will be the winner yet. Superconducting qubits is currently most widely adopted and there the main requirement is sample cleanliness. The critical dimensions are huge vs typical semi chips. For example 1 typical SC qubit can be mm in size, and consist of a huge superconducting aluminum LC resonator (with a giant coplanar capacitor) coupled to a Josephson junction built around an aluminum oxide tunnel barrier, all on a sapphire substrate. One problem they are trying to solve right now is that any defect in the material, including the polymer resist residue, will cause decoherece.
There are many different platforms to build quantum technologies and they all have very different requirements. It isn’t clear which platform will be the winner yet. Superconducting qubits is currently most widely adopted and there the main requirement is sample cleanliness. The critical dimensions are huge vs typical semi chips. For example 1 typical SC qubit can be mm in size, and consist of a huge superconducting aluminum LC resonator (with a giant coplanar capacitor) coupled to a Josephson junction built around an aluminum oxide tunnel barrier, all on a sapphire substrate. One problem they are trying to solve right now is that any defect in the material, including the polymer resist residue, will cause decoherece.
I suppose the low volume and small chip size justifies the use of e-beam (short write time), but I imagine they could alternatively use i-line (365 nm) lithography with spacers for higher volumes.
I believe materials compatibility will be a challenge. If JJ based qubits triumph, it's less of a problem because they are aluminum based, but other approaches use all sorts of metals that are not very CMOS compatible and would pose a serious contamination threat to any modern facility.
