HPC-Quantum The energy transition is also a paradigm shift in computing
Italian energy company Eni and neutral-atom quantum computing startup Pasqal are collaborating to develop quantum-based high-performance computing (HPC) solutions to accelerate the green energy transition.
Quantum computing uses the principles of quantum physics to solve problems that are beyond the capabilities of traditional computers. Paris-based Pasqal has created quantum computers using ordered neutral atoms arranged in 2D and 3D arrays. Through its full-stack approach, Pasqal is creating quantum algorithms to improve current HPC processes for a range of energy applications that Eni is working on in its value chain (upstream, downstream, chemicals and renewable energy).
Since 2021, Eni has invested in Pasqal since 2021 through its venture capital arm, Eni Next, which specializes in investing in startups that address the transition to a zero-carbon future for energy. To date, Pasqal has raised over €40 million from various sources.
Today, HPC is an important tool for solving complex problems in science, finance and medicine. This includes climate system modeling, earthquake prediction, vaccine development, materials design, drug development, inventory management and image recognition, just to name a few applications where HPC has made significant contributions.
In recent years, incremental performance improvements have been achieved through the introduction of memory-sharing solutions, the availability of high-performance processors (such as GPUs and FPGAs), and the widespread use of parallel architectures (such as clustered servers widely used in the cloud) to speed up the execution time of the most complex algorithms, including artificial intelligence.
But while today's HPC systems run more than a million times faster than the fastest desktop, laptop, or server systems, there are some problems that even the most advanced supercomputers cannot solve in a feasible timeframe. Quantum computing promises performance unattainable by traditional systems, effectively solving highly complex computational problems that take too long on traditional computers.
This includes modeling the evolution of large-scale systems. For example, weather analysis models must determine the future behavior of atmospheric factors over vast areas covering tens or hundreds of square kilometers. Such models are based on systems of differential equations that require solving for many variables. Today's modeling systems are designed to produce accurate and timely forecast data, but their limitations mean that forecasters must settle for a level of reliability that typically does not exceed 50%.
Similar problems of computational complexity include earthquake prediction, oil field fluid flow analysis, molecular systems analysis (DNA sequencing), and automated trading systems. Quantum computers can execute quantum algorithms capable of solving these and other problems.
02Eni's High Performance Computing Journey
Eni invests in digital technologies to improve its operational efficiency, enable new business and support a net reduction in global greenhouse gas (GHG) emissions, with the goal of providing fully decarbonized products and services by 2050. The company's green data center in Ferrera Erbognone, Italy, has an HPC4 supercomputer, installed in 2018, and a subsequent upgrade, the HPC5 (Figure 1), which entered the Global Supercomputing 500 ranking in June 2020. the HPC5 also debuted on the GREEN500 special list in June 2020, sixth in the world, first in Europe , tied for first overall in industrial computing for the most energy-efficient supercomputer, a position it holds until June 2022.
Eni says it has designed one of the most sustainable and energy-efficient data centers in the world. The center uses less electricity to power its computers than competing facilities, according to the company.
Figure 1 HPC5 by Eni
The Inertial Wave Energy Converter (ISWEC) system, developed in collaboration with the Politecnico di Torino, is one of the applications realized by HPC4 and HPC5. ISWEC generates electricity from wave energy, which is considered to be the largest untapped renewable energy source on the planet. The converted energy can be used immediately in offshore power plants or fed directly into the grid to provide power to coastal communities. The first pilot plant is already in operation in the high seas pilot area off the coast of La Verna, Italy (Figure 2).
Figure 2 ISWEC's pilot plant is now operational in the high seas near La Verne
The Green Data Center supercomputers are also being used to run new algorithms for magnetic confinement fusion research to support the development of superconducting magnets and plasma research. These computational tools are also being used to generate theoretical models of molecules and photoactive polymers, which are at the heart of Eni's solar energy capture technology.
Leveraging Pasqal's proprietary algorithms will allow Eni to accelerate research in these and other areas and unlock new capabilities, the companies said.
03Pasqal's Quantum Processor
Pasqal has developed a quantum processor (QPU) that is based on a neutral atomic device architecture and is particularly well suited for simulations. Simulation using synthetic quantum systems is a promising tool for solving complex non-deterministic polynomial-time hard (NP-Hard) problems, which are usually not solved by traditional numerical methods.
Quantum simulations are the most promising use of Pasqal's QPU, where the QPU is used to understand the quantum system of interest. Neutral-atom quantum processors will benefit scientific discovery and industrial applications, such as the design of novel materials for energy storage and transport or chemical calculations for drug discovery.
In Pasqal's quantum processor, the quantum bits are atoms; atoms are manipulated one after another to store and encode quantum information. pasqal says this approach has allowed the company to reach an unprecedented system size with more than 100 atomic quantum registers. A large number of interacting quantum particles will be able to simulate the dynamics of many-body quantum systems, which is beyond the capabilities of conventional techniques.
Figure 3 A 14 × 14 filled array of atoms corresponding to 196 quantum bits
Pasqal cites several benefits of its approach. First, it does not require cooling and can work at room temperature. Second, because each object is naturally occurring, they are all structurally identical. Third, the quantum computer developed by Pasqal consumes as much energy as four hair dryers.
The ability to control hundreds of quantum bits shows that this system is very scalable. The company says the technology is already capable of controlling up to 200 quantum bits. Figure 3 shows 196 quantum bits as a 14 × 14 filled array of atoms.
04Cooperation, not competition
Quantum computers are powerful, but they also have their drawbacks. First, quantum computers are difficult to use, require a very controlled setup to operate, and must cope with the loss of quantum states due to "decoherence", which can produce strange results. Quantum computers are also rare, expensive, and less efficient than traditional computers for most tasks.
Nevertheless, many of these problems can be offset by combining quantum computers with traditional computers. As the Finnish research institute VTT chose to connect a small quantum computer to Europe's most powerful classical supercomputer, HELMI, a 5-bit superconducting quantum computer developed by VTT, to LUMI, one of the top three supercomputers in the world in terms of computing power.
In this architecture, researchers can create a hybrid algorithm that lets the traditional supercomputer LUMI handle the parts it does best, while giving anything that could benefit from quantum computing to HELMI. then, LUMI can integrate the results of HELMI's quantum calculations, perform any additional calculations necessary, send even more calculations to HELMI, and send the complete results back to the researcher.
