MODULAR QUANTUM COMPUTING SYSTEM FOR DISTRIBUTED QUANTUM COMPUTATION VIA QUANTUM ENTANGLEMENT
A modular quantum computing system that enables distributed quantum computation across multiple quantum processing units (QPUs) that are remotely connected using a quantum entanglement network is disclosed. In order to execute a quantum circuit across multiple QPUs, any quantum state pertaining to any given multi-qubit gate of the quantum circuit may be teleported between two respective QPUs, such that an overall quantum compute capacity for executing the quantum circuit is expanded. Based on a number of QPUs that are allocated for executing the quantum circuit, buffers of established, pairwise quantum entanglement instances between respective sets of the allocated QPUs may be prepared and subsequently maintained prior to and during execution of the quantum circuit, in order to limit potential latency due to use of a quantum entanglement network within the execution of a given quantum circuit.
1 . A modular quantum computing system, comprising:
a quantum entanglement network subsystem configured to remotely connect separate quantum processing units (QPUs) using optical communications links;
a first QPU comprising:
a first set of physical qubits, designated for quantum computation operations; and
a second set of physical qubits, designated for quantum entanglement operations; and
a second QPU comprising:
a third set of physical qubits, designated for quantum computation operations; and
a fourth set of physical qubits, designated for quantum entanglement operations,
wherein, to execute a given multi-qubit gate of a given quantum circuit between a respective one of the first set of physical qubits and a respective one of the third set of physical qubits, the quantum entanglement network subsystem is further configured to teleport a quantum state of a respective one of the second set of physical qubits to a respective one of the fourth set of physical qubits.
2 . The modular quantum computing system of claim 1 , wherein:
the quantum entanglement network subsystem comprises:
a first quantum repeater, locally connected to the first QPU, wherein the first quantum repeater comprises a first set of quantum memories; and
a second quantum repeater, locally connected to the second QPU, wherein the second quantum repeater comprises a second set of quantum memories; and
the quantum entanglement network subsystem is further configured to:
establish one or more pairwise quantum entanglement instances, using one or more of the optical communications links, with respective ones of the first set of quantum memories of the first quantum repeater and respective other ones of the second set of quantum memories of the second quantum repeater.
3 . The modular quantum computing system of claim 2 , wherein:
the first quantum repeater further comprises a first optical switchboard;
to teleport the quantum state of the respective one of the second set of physical qubits to the respective one of the fourth set of physical qubits, the first optical switchboard is configured to perform a Bell state measurement between the respective one of the second set of physical qubits and a given quantum memory of the first set of quantum memories;
the second quantum repeater further comprises a second optical switchboard; and
to teleport the quantum state, the second optical switchboard is configured to perform a Bell state measurement between the respective one of the fourth set of physical qubits and another given quantum memory of the second set of quantum memories.
4 . The modular quantum computing system of claim 3 , wherein:
the first quantum repeater further comprises an optical transducer configured to enable the first optical switchboard to interface with signals obtained from the second set of physical qubits in the first QPU.
5 . The modular quantum computing system of claim 1 , wherein to execute the given multi-qubit gate of the given quantum circuit between the respective one of the first set of physical qubits and the respective one of the third set of physical qubits, the first QPU is further configured to:
execute one or more SWAP gate operations between the respective one of the first set of physical qubits, one or more other physical qubits of the first set of physical qubits, and the respective one of the second set of physical qubits.
6 . The modular quantum computing system of claim 1 , wherein the first QPU is further configured to execute one or more additional gates of the given quantum circuit between respective other ones of the first set of physical qubits.
7 . A system, comprising:
one or more classical computing devices of a service provider network configured to implement an elastic quantum computing service configured to orchestrate execution of quantum circuits using a plurality of quantum processing units (QPUs) made accessible via the service provider network, wherein, to implement the elastic quantum computing service, the one or more classical computing devices are further configured to:
allocate a number of QPUs, of the plurality of QPUs, to be used in executing a given quantum circuit; and
determine gate scheduling instructions to be applied during execution of the given quantum circuit across the allocated number of QPUs, wherein, to determine the gate scheduling instructions, the one or more classical computing devices are further configured to schedule a multi-qubit gate to be executed using a physical qubit of a first QPU and a physical qubit of a second QPU of the allocated number of QPUs; and
a quantum entanglement network comprising a plurality of quantum repeaters locally connected to respective ones of the plurality of QPUs,
wherein to execute the multi-qubit gate using the physical qubit of the first QPU and the physical qubit of the second QPU, the quantum entanglement network is configured to cause quantum entanglement to be generated between a first quantum repeater of the plurality of quantum repeaters, locally connected to the first QPU, and a second quantum repeater of the plurality of quantum repeaters, locally connected to the second QPU.
8 . The system of claim 7 , wherein to allocate the number of QPUs to be used in executing the given quantum circuit, the one or more classical computing devices implementing the elastic quantum computing service are further configured to:
determine a minimum number of physical qubits that are to be used to execute the given quantum circuit based, at least in part, on a given compiled version of the given quantum circuit;
determine one or more combinations of QPUs of the plurality of QPUs that result in at least the minimum number of physical qubits; and
allocate the number of QPUs to be used in executing the given quantum circuit based, at least in part, on the one or more combinations of QPUs.
9 . The system of claim 8 , wherein to determine the one or more combinations of QPUs of the plurality of QPUs that result in at least the minimum number of physical qubits, the one or more classical computing devices implementing the elastic quantum computing service are further configured to:
determine QPUs of the plurality of QPUs that are currently allocated, or are scheduled to be allocated, for use in executing other quantum circuits; and
determine the one or more combinations of QPUs of the plurality of QPUs that result in at least the minimum number of physical qubits based, at least in part, on the determination of the QPUs of the plurality of QPUs that are currently allocated, or are scheduled to be allocated, for use in executing the other quantum circuits.
10 . The system of claim 7 , wherein the one or more classical computing devices implementing the elastic quantum computing service are further configured to:
generate quantum entanglement instructions to be provided to the quantum entanglement network prior to the execution of the given quantum circuit across the allocated number of QPUs, wherein the quantum entanglement instructions indicate one or more pairwise quantum entanglement instances that are to be established between respective ones of the plurality of quantum repeaters based, at least in part, on the determined gate scheduling instructions.
11 . The system of claim 10 , wherein the quantum entanglement network is configured to establish the one or more pairwise quantum entanglement instances between the respective ones of the plurality of quantum repeaters based, at least in part, on the provided quantum entanglement instructions.
12 . The system of claim 11 , wherein:
the quantum entanglement network is further configured to maintain a buffer of the established one or more pairwise quantum entanglement instances such that a rate of establishing the one or more pairwise quantum entanglement instances is higher than a rate of decay of the one or more pairwise quantum entanglement instances; and
the rate of decay of the one or more pairwise quantum entanglement instances is based, at least in part, on coherence times of qubits within respective quantum memory locations of the respective ones of the plurality of quantum repeaters.
13 . The system of claim 7 , wherein:
to determine the gate scheduling instructions of the given quantum circuit across the allocated number of QPUs, the one or more classical computing devices are further configured to schedule a subsequent multi-qubit gate to be executed using an additional physical qubit of the first QPU and a physical qubit of a third QPU of the allocated number of QPUs; and
the subsequent multi-qubit gate is dependent upon, at least in part, an output of the multi-qubit gate to be executed using the physical qubit of the first QPU and the physical qubit of the second QPU.
14 . The system of claim 7 , wherein:
the system further comprises a third quantum repeater, configured to establish one or more pairwise quantum entanglement instances with the first quantum repeater, and one or more additional pairwise quantum entanglement instances with the second quantum repeater; and
to execute the multi-qubit gate using the physical qubit of the first QPU and the physical qubit of the second QPU, the quantum entanglement network is configured to cause distributed quantum entanglement to be generated between the first quantum repeater and the third quantum repeater, and between the third quantum repeater and the second quantum repeater.
15 . The system of claim 7 , wherein:
the first QPU is located at a premises within the service provider network;
the first quantum repeater comprises:
a set of quantum memories; and
an optical switchboard; and
to execute the multi-qubit gate using the physical qubit of the first QPU and the physical qubit of the second QPU, the optical switchboard is configured to perform a Bell state measurement between a given quantum memory of the set of quantum memories and another physical qubit of the first QPU, designated for quantum entanglement operations.
16 . The system of claim 15 , wherein:
the second QPU is located at the premises within the service provider network;
the second quantum repeater comprises:
another set of quantum memories; and
another optical switchboard; and
to execute the multi-qubit gate using the physical qubit of the first QPU and the physical qubit of the second QPU, the other optical switchboard is configured to perform another Bell state measurement between another given quantum memory of the other set of quantum memories and another physical qubit of the second QPU, designated for quantum entanglement operations, wherein the other given quantum memory of the other set of quantum memories within the second QPU corresponds to an established, pairwise quantum entanglement instance with the given quantum memory of the set of quantum memories within the first QPU.
17 . A method, comprising:
receiving a request from a customer of an elastic quantum computing service to execute a quantum circuit using quantum computing resources of the elastic quantum computing service;
allocating a number of quantum processing units (QPUs), of a plurality of QPUs made available by the elastic quantum computing service, for use in executing the quantum circuit, wherein the allocated QPUs are remotely connected using quantum repeaters of a quantum entanglement network;
executing the quantum circuit using the allocated QPUs, wherein said executing the quantum circuit comprises:
executing a given multi-qubit gate of the quantum circuit between a physical qubit of a first QPU of the allocated QPUs and a physical qubit of a second QPU of the allocated QPUs, wherein said executing the given multi-qubit gate comprises teleporting a quantum state, pertaining to the given multi-qubit gate, between another physical qubit of the first QPU, designated for quantum entanglement operations, and another physical qubit of the second QPU, designated for quantum entanglement operations; and
providing execution results of the quantum circuit to the customer.
18 . The method of claim 17 , wherein said executing the given multi-qubit gate of the quantum circuit between the physical qubit of the first QPU and the physical qubit of the second QPU further comprises:
executing, prior to said teleporting the quantum state, one or more SWAP gate operations between the physical qubit of the first QPU and the other physical qubit of the first QPU, designated for quantum entanglement operations.
19 . The method of claim 17 , wherein said executing the quantum circuit using the allocated QPUs further comprises:
responsive to said executing the given multi-qubit gate of the quantum circuit between the physical qubit of the first QPU and the physical qubit of the second QPU,
executing one or more subsequent multi-qubit gates of the quantum circuit using one or more of the allocated QPUs, wherein the one or more subsequent multi-qubit gates are dependent upon, at least in part, an output of the multi-qubit gate executed between the physical qubit of the first QPU and the physical qubit of the second QPU.
20 . The method of claim 17 , wherein said allocating the number of QPUs for use in executing the quantum circuit comprises:
determining a minimum number of physical qubits that are to be used to execute the quantum circuit based, at least in part, on a given compiled version of the quantum circuit;
determining one or more combinations of QPUs of the plurality of QPUs that result in at least the minimum number of physical qubits; and
allocating the number of QPUs based, at least in part, on the one or more combinations of QPUs.