IonQ has unveiled the foundation for future multi-core quantum processors with configurable qubit configurations. The processor is made of quartz glass and promises to easily manufacture and operate quantum computing - no cryogenics, right down to rack-mountable computing in conventional computing rooms.
Until now, optical ion trap quantum processors have been manufactured using silicon substrates for systems on a chip. IonQ also used this technology, but it began producing new processors on silica glass substrates. This technology is widely used for the production of microfluidic microcircuits for the tasks of analytical chemistry and medicine, therefore, IonQ has no problems with factory facilities and will not have problems - it all works and is quite inexpensive.
In addition to production efficiency, the transition to glass gives additional and more important effects - this is the absence of random electromagnetic fields, which are possible in silicon, and transparency for lasers in a wide optical range. The first gives a low level of interference and increases the accuracy of calculations with qubits, and the second allows the laser beams to penetrate into the chip with the least loss and reach the ions (qubits) held in the trap in order to do something useful with them.
Glass also simplifies the creation of optical connections for pairing several quantum processors into one computing node, and this is a relatively simple scaling of quantum systems, which is what everyone involved is dreaming of.
To prove the effectiveness of the new proposal, IonQ introduced a 64-qubit quartz glass quantum processor with sputtering (deposition) of metal components. The processor is a linear ion trap in which ions are arranged in a chain. The IonQ processor operates 4 strings of 16 ions each. However, 4 ions in each chain are used for technical needs, if I may say so: they participate in the so-called collisional cooling operations designed to stabilize the computational ions while moving through the trap. Thus, the number of computational qubits is reduced to 48, but in fact it is even less - 32, if we take into account error correction.
In a linear trap, chains can be mated, linking up to 32 qubits. The laser beams incident on the chip can control only 16 qubits at a time (one chain at a time). These lasers define both the quantum states of individual qubits and link pairs. By conjugating the chains, it is possible to sequentially link all the qubits in the processor, which is quite simple compared to superconducting qubits, as in the IBM or Google system.
IonQ sees the most valuable thing in the new development as the ease of scaling both within the chains within the processor, and from the point of view of connecting multiple processors. A sharp increase in productivity occurs already at the stage of adding each new ion (qubit) in the chain. In general, IonQ expects to increase the number of qubits in the proposed architecture to a three-digit value. And this event is not far off. IonQ systems are very close to commercial production and may become available in the coming years.
0 Comments:
Post a Comment
Your comment and facebook share will be appreciated