In recent years, quantum computing has experienced rapid development, with companies like Amazon, IBM, and traditional silicon manufacturers working hard to achieve effective error correction. There is a strong consensus that in order to tackle the most useful problems with a quantum computer, it must be capable of error correction. However, there is no consensus on which technology will allow us to achieve this.
Recently, three articles have been published addressing different aspects of quantum computing technology. An international team including startup Diraq has demonstrated that a silicon quantum dot processor can perform well at the relatively warm temperature of 1 Kelvin, compared to the usual milliKelvin at which these processors normally operate. This suggests that the chips can tolerate reasonable operating temperatures, meaning control circuitry can be used on the chip without causing issues.
On the other hand, IBM researchers have described a new error correction model for use with superconducting qubits, called transmons. Using a scheme called “low-density parity-check codes” (LDPC), simulations show they are able to handle a dozen logical qubits using only 288 physical qubits, much fewer than would be needed for a useful surface code. However, to implement these ideas in hardware, IBM will have to roughly double the number of qubit-to-qubit connections of their existing configurations and build chips with longer-range connections.
Amazon is also developing their own efforts to develop quantum computers based on transmons, but they are using them in different ways to try to reduce error rates. In a paper published last week, a team of Amazon researchers and academics describe error correction using what they call a dual-rail transmon. The result is a qubit where the inherent error rate is well below what is needed for error correction schemes to work. However, error checking is relatively slow compared to other operations, which can slow down calculations.
While some of these advancements may be significant on their own, not all of them are compatible as transmons and quantum dots are completely different physical systems. However, these efforts show that even if one is fundamentally skeptical about the prospects of quantum computing, there is interesting work being done to try to provide the necessary components for things to work.
It is still unclear how much of this will be incorporated into future quantum computing efforts, but it does demonstrate that people working to advance the field have not run out of ideas. In the coming years, we probably will not have a clearer picture of what is likely to work, but there will be a lot of interesting research and development work between now and then, some of which may represent key milestones in the development of quantum computing.