IBM Runs Quantum Hardware Optimally to Achieve a New Performance Milestone

IBM has upgraded one of its newest 27-qubit client-deployed systems to achieve a Quantum Volume 64. Congratulations to IBM for advancing quantum computer technology that matches the best quantum volume metric for a quantum computer. Honeywell also claims a Quantum Volume 64 quantum computer. The best a regular non-quantum computer working on the type of problems has achieved the equivalent of a Quantum volume of 4096. This is explained below.

IBM used new techniques and improvements that used knowledge of the hardware to optimally run the Quantum Volume circuits. These hardware-aware methods are extensible and will improve any quantum circuit run on any IBM Quantum system, resulting in improvements to the experiments and applications which users can explore. These techniques will be available in upcoming releases and improvements to the IBM Cloud software services and the cross-platform open-source software development kit (SDK) Qiskit.

IBM Quantum Highlights

*IBM has reached Quantum Volume 64 on a 27-qubit system deployed within the IBM Q Network []
28 quantum computing systems deployed on the IBM Cloud over the last four years with eight systems boasting a Quantum Volume of 32
* The IBM Q Network has 115 client, government, startup, partner, and university members
* 250,000+ registered users of the IBM Quantum Experience []
* Users routinely execute more than 1 Billion hardware circuits per day on IBM Quantum systems on the IBM Cloud
* Researchers have published 250+ papers based on work on IBM Quantum systems

Quantum Volume Background

Most people have no sense of what the quantum volume number means. For most people, quantum volume is as meaningless as a metric as Power Level in Dragon Ball Z.

Quantum volume measures the number of qubits, stability of the qubits, connectedness and several other characteristics that affect what can be solved.

The goal of creating quantum volume was to try to match up with the LINPACK benchmark for regular supercomputers. The LINPACK Benchmarks are a measure of a system’s floating-point computing power. They measure how fast a classical computer solves a dense n by n system of linear equations Ax = b, which is a common task in engineering.

Different researchers compared NASA’s Electra supercomputer, which is primarily powered by Intel Skylake CPUs, with ORNL’s Summit supercomputer, which is primarily powered by NVIDIA Volta GPUs. Tens of petaflop classical computers solved a 7X7 circuit which is a Quantum volume of 12.

In 2018, Alibaba and the University of Michigan published “Classical Simulation of Intermediate-Size Quantum Circuits”.

Computing a single amplitude of an 8 × 8 qubit circuit with depth 40 was previously beyond the reach of supercomputers. Their algorithm can compute this within 2 minutes using a small portion (≈ 14% of the nodes) of the cluster.

They successfully simulating quantum supremacy circuits of sizes
9×9×40 (QV 512),
10×10×35 (QV 1024),
11 × 11 × 31 (QV2048), and
12 × 12 × 27 (QV4096).

Regular (non-quantum) systems achieved a Quantum volume of 4096.

They did that using 131072 processors and 1 petabyte of memory. They give evidence that noisy random circuits with realistic physical parameters may be simulated classically. This suggests that either harder circuits or error-correction may be vital for achieving quantum supremacy from random circuit sampling.

In June, Honeywell released a quantum computer that Honeywell claims has a quantum volume of 64.

SOURCES- IBM, Arvix, Quantum Computer Reports, Honeywell
Written By Brian Wang,

3 thoughts on “IBM Runs Quantum Hardware Optimally to Achieve a New Performance Milestone”

  1. Yes, it means that, to the degree that QV is a good metric. But it has some flaws, so it’s possible two computers would have the same QV but differ quite a bit in how well they can solve problems.

    A more meaningful measure might be how many logical qubits you can build, which use quantum error correction so they can go for billions of operations in a row, rather than dozens. We can’t yet build enough qubits to get even 1 logical qubit. But maybe someday we will.

  2. At what QV we will have game-changing application? Somewhere I read 1000, but that sounded out of thin air. Does the simulation of 4096 QV means that a classical computer may have the same applications a quantum computer could have with 4096 QV?

  3. It’s impressive that they managed to keep 27 qubits coherent long enough to do 30 operations before they failed. They describe that as a quantum volume of 64.

    For comparison: to break cryptography with RSA 2048, you’ll need to keep 4000 logical qubits coherent long enough to do 9 billion operations.

    And you might need hundreds of thousands of actual qubits to get the 4000 logical qubits.

    There are some criticisms of using quantum volume as a metric:

    The IBM paper:

Comments are closed.