First of all, it's important to distinguish between a supercomputer and a quantum computer. While the objectives are the same, the main difference lies in the way they process and solve problems.

A classical computer is driven by one or more processors at its heart, and the speed with which it processes and analyzes data depends on the power of these processors. A supercomputer, like those from IBM or Atos, uses conventional processors (GPUs) to perform a very large number of calculations simultaneously. They are particularly effective at solving so-called "deterministic" problems in parallel, such as weather simulation, fluid dynamics or the analysis of large databases. These machines are optimized to maximize the speed and number of calculations per second (measured in floating-point operations per second (FLOPS)).

To increase this power, increasingly tiny components are used, reaching sub-nanometer dimensions, where the rules of quantum physics take over from those of classical physics. This is where quantum computing comes in. Unlike classical computers, which store information in the form of 0s or 1s (bits), a quantum computer can record and process these two states simultaneously (Qubits) thanks to a phenomenon known as superposition, which dramatically multiplies its computing power.

Source : CBInsight

As the technology is still in the R&D stage, there is no single way of making a quantum computer. Today, six Qubit methods are in use at different stages of maturity:

  • Superconducting Qubits : These operate at temperatures close to absolute zero, enabling precise control of quantum states. Although this technology is among the most advanced, it requires extensive cooling systems.
  • Ion qubits: These qubits use laser-controlled ions in a vacuum. However, they are not as fast as superconducting qubits, and large-scale implementation is problematic.
  • Photon qubits: These exploit elementary particles of light to perform quantum operations, manipulating them via optical devices. However, the interaction between photon qubits remains a challenge for large-scale calculations to avoid errors.
  • Neutral-atom qubits: Individual, uncharged atoms with an equal number of electrons and protons can store information in different energy levels. These atoms can be arranged in very dense arrays, facilitating their use for large-scale systems, but interaction problems remain.
  • Nuclear or electron spin Qubits: This method is based on the spin of electrons or atomic nuclei, manipulated by magnetic fields. Mainly explored in academic research, they offer prolonged coherence, favorable to computational stability, but their precise control and scaling pose technical challenges.
  • Topological Qubits: Still at the theoretical stage, this method is still poorly documented, but is still used by Microsoft.

So, let's take a look at the listed companies positioned in the supercomputer and quantum computer segment.

Pure-players in the quantum field

  • IonQ focuses on trapped-ion quantum computing, with systems available via all major cloud providers. It is the only company whose quantum systems are available on Amazon Braket, Microsoft Azure and Google Cloud. It was the first pure-player in QC.
  • Rigetti Computing develops superconducting quantum processors and has recently created a modular qubit architecture. The company is also developing a cloud-based platform called Forest that enables programmers to write quantum algorithms.
  • Quantum Computing specializes in room-temperature quantum systems and marketable applications in the fields of logistics, bioinformatics, life and physical sciences, quantitative finance and electronic design automation.
  • D-Wave Quantum is a pioneer in "quantum entanglement" technology, and provides a quantum cloud system for its customers.

Tech giants with quantum programs

  • IBM is a historic giant in supercomputing. The company leads the way in quantum computing (QC). They have launched the IBM Q Experience, enabling over 100 customers to access quantum resources via the cloud. Their 127-qubit Eagle quantum processor is a leading QC option used by leading companies such as Goldman Sachs and Boeing. The company is betting heavily on this segment.
  • Google (Alphabet) has made headlines in the quantum field in recent years, particularly after the publication of its groundbreaking paper in Nature, claiming that its Sycamore quantum processor was the first to achieve "quantum supremacy". However, like the other major companies mentioned, Alphabet is a highly diversified conglomerate, which means that the contribution of quantum computing to Alphabet's overall valuation is likely to be modest.
  • Microsoft, which focuses on topological qubits and offers quantum services via Azure Quantum, is actively developing quantum software, including the Q# SDK (pronounced "Q sharp"). They also invest in PsiQuantum via their M12 venture capital arm. However, the absence of any mention of quantum in their annual financial report highlights a potentially limited impact on their stock market value in the short term.
  • Intel is a serious candidate for significant involvement in quantum computing. In 2019, Intel announced Horse Ridge, the Tunnel Falls quantum chip, and hopes to use it to become a major player in QC.
  • Nvidia and its GPUs are now also being used for deep learning in AI. The giant is playing a key role in advanced computing and is getting into QC with cuQuantum. Although quantum is still a minor field for Nvidia, their expertise could foster future success in this sector, making a long-term investment position attractive.
  • Amazon, like Microsoft, offers a cloud-based quantum platform within AWS, called Braket. This platform provides access to systems from D-Wave, Rigetti and IonQ. However, Amazon is a large company with many interests, and the term "quantum" is not often highlighted in its presentation documents.
  • RTX Corp is collaborating with IBM on quantum research for aerospace, defense and intelligence applications.
  • Tencent is the largest Chinese company still involved in quantum computer research. Indeed, Chinese tech giants Alibaba and Baidu pulled out of the quantum computing race after closing their research units earlier this year, without giving any specific explanations.

Industrial companies

  • Honeywell: the company has spun off its Honeywell Quantum Solutions (HQS) division to form Quantinuum, with the announced intention of taking it public. In the meantime, investing directly in Honeywell provides access to the quantum computing sector. Even after Quantinuum's IPO, Honeywell is expected to retain a significant stake.

Supercomputers

Most listed companies in the quantum computing segment also own supercomputers. However, these are not the same companies that are leading the way in terms of performance and innovation. We'll use the Top 500 supercomputer performance rankings here for illustrative purposes.

  • Hewlett Packard Enterprise is the current market leader. With its Frontier computer, HPE has the most powerful supercomputer since 2022, ahead of Intel (2nd), Microsoft (3rd), IBM (9th) and Nvidia (10th). The company also has four supercomputers in the top 10 of the annual ranking of the 500 most powerful supercomputers.
  • Fujitsu, the Japanese multinational conglomerate, has a long history in the supercomputer sector. With its Fugaku computer, the group holds the fourth most powerful computer in the top 500. The company's supercomputers are used in a wide range of fields, such as industrial simulation, weather forecasting and scientific research.
  • Lenovo, China's leading PC manufacturer, has increased its share of the supercomputer market.
  • AMD, Nvidia's rival microprocessor manufacturer, has become a major force in the supercomputer market. Supercomputing systems have become dependent on AMD's EPYC processors and Radeon Instinct gas pedals for their competitive performance and energy efficiency.
  • Atos, under the name Eviden (cloud computing, big data and security part of the group), is one of Europe's only supercomputer manufacturers. Its Leonardo (7th) and MareNostrum (8th) computers are the most powerful in Europe, and were acquired from Bull in 2014. There are currently rumors of a split between Eviden and Atos.
  • Orange, as part of its partnership with HP, will be in charge of the software side, while its partner will supply the hardware. This will enable the company to maintain its national independence, since computers are used in crucial fields such as defense and nuclear power. This decision may not be final. Options will be re-evaluated at computer changeover time (generally every two years).

However, pure players such as ParTec, a German start-up, and Canaan, an American small cap, are dedicated to developing and supplying supercomputers and quantum computers to assist their customers in training AI models.

Only one ETF is indirectly exposed to supercomputers and quantum computers:
  • Defiance Quantum ETF: this tracker seeks to track the performance of the BlueStar Quantum Computing and Machine Learning index, which in turn replicates the performance of the largest and most liquid companies in the global quantum computing and machine learning sectors. The top five positions are IonQ (5.16%), MicroStrategy (3.30%), D-Wave Quantum (2.48%), Rigetti Computing (2.16%) and Coherent Corp (1.90%). With $297 million in assets under management, fees are 0.40% and year-to-date performance exceeds 18%.