Fujitsu develops quantum/HPC hybrid computing technology to optimize solution brokering for customers : Fujitsu Global

Fujitsu develops quantum/HPC hybrid computing technology to optimize solution brokering for customers : Fujitsu Global

New technologies delivers much easier obtain to next-generation computing technologies to non-computing gurus

Fujitsu Minimal

Tokyo, November 8, 2022

Fujitsu right now announced the growth of quantum/HPC hybrid computing engineering to improve workload variety for prospects. The new AI-based computer software, which serves as a precursor to a future personal computer workload broker technologies, quickly selects from distinctive following-era computing platforms (1) to present the exceptional answer to customers’ complications dependent on parameters which include calculation time, calculation accuracy, and fees.

The new technologies leverages the electricity of Fujitsu’s environment leading-class 39-qubit quantum simulator and the “FUJITSU Supercomputer PRIMEHPC Forex 700,” equipped with the same A64Forex CPU that powers supercomputer Fugaku, combining and deciding on the the best possible calculation strategy for customers’ quantum chemical calculation (2) troubles. Crucially, the technologies makes it possible for people devoid of specialized expertise to use quantum simulators and HPC engineering to provide successful alternatives to true-globe issues.

Entire world-foremost highly developed computing systems for alternatives to 21st century challenges

Fujitsu has a long monitor file in creating and supplying entire world-primary computing systems, exemplified by its progress and attempts to commercialize the supercomputer Fugaku, subsequent-era quantum computing, as perfectly as technologies that bridge the hole among regular and quantum, like the Electronic Annealer and quantum simulators.

To promote the practical enhancement of new quantum algorithms, Fujitsu declared the world’s quickest quantum simulator in March 2022, initiating joint exploration in the industry of elements science with Fujifilm Company (3) and Tokyo Electron Ltd. (4).

Easy accessibility to future-generation computing systems for non-computing specialists

Fujitsu is at this time doing the job to create a computing workload broker—a software program technology that employs AI to automatically pick the correct assets from sophisticated computing systems and methods, which include big-scale cloud-based HPC, quantum computing, Electronic Annealer, as nicely as quantum simulators to solve customers’ complex difficulties.

Fujitsu envisions the foreseeable future computing workload broker as supporting both of those the automation and the optimization of calculation processing of numerous hardware products and the optimization of calculation distribution for unique platforms. In this way, the computing workload broker will empower superior-velocity calculations and options in accordance to customers’ unique needs.

Figure 1: Picture of the computing workload broker

Outline quantum /HPC hybrid computing technological know-how

To notice its eyesight for the computing workload broker, Fujitsu has now produced a quantum/HPC hybrid calculation technological innovation for resolving quantum chemical problems to allow calculation-based mostly clarification of the attributes of products applied in drug discovery and new substance advancement. The new know-how, which serves as a precursor to the computing workload broker, permits large precision calculations at significant speed by quickly and optimally combining two sorts of desktops: a quantum simulator and HPC. The capabilities of the new know-how are as follows.

Figure 2: Comparison of traditional technological innovation and freshly made technological innovation

1. Quantum/HPC algorithm discrimination technologies

Algorithms utilized in quantum chemical calculations involve recurring calculations right up until a really accurate solution is attained. As the distance concerning atoms improvements, determining whether quantum or HPC algorithms offer the optimal option to a issue represents a challenging problem for present technologies. To this finish, Fujitsu produced a remedy to identify the accurate algorithm for a problem by examining how the algorithm converges on molecules. For problems where by classical algorithms do not deliver adequate results, the technological innovation detects a unique pattern in the convergence condition right up until the algorithm calculates a alternative, so that the optimum algorithm can be identified by doing pre-processing on the dilemma experimentally employing HPC algorithms.

2. Computation time estimation engineering

Precise estimation of the convergence of the many molecular structures represented a complicated activity in quantum chemical calculations, and it proved complicated to estimate the time and value demanded to acquire hugely correct solutions in progress. To address this concern, Fujitsu manufactured an AI design equipped to estimate calculation amounts, expenditures and time in advance by utilizing an adaptive AI technological know-how made by Fujitsu to analyze the romantic relationship involving the molecular construction, the iterative calculation of the algorithm and the calculation time.

3. Optimal manage technological innovation based mostly on time, charge and precision

In purchase to help consumers to complete calculations at ideal expenditures and in ideal time, Fujitsu formulated a technological innovation that optimizes the overall performance of quantum chemical calculations. The know-how takes into account both of those the calculation time and charge estimated by the quantum/HPC algorithm discrimination technological innovation and the computation time estimation technological innovation as well as the usage of calculation methods. In this way, the know-how enables consumers to resolve quantum chemical calculation complications in a way that best fulfills their desires without the need of the will need to decide on from unique computational means by themselves.

Figure 3: Overview of quantum-HPC hybrid computing technological know-how

Future Options

Shifting forward, Fujitsu will confirm the success of this know-how and more build it, with the aim of creating a new computing workload broker technology in the field of quantum chemical calculations by fiscal 2023. Fujitsu will further continue on to develop computing platforms that can be applied by anyone with no the have to have for abilities by introducing providers to Fujitsu Computing as a Service (hereinafter “CaaS”), Fujitsu’s assistance portfolio to deliver consumers entry to planet-main computing technologies through the public cloud.


Associated Links

Fujitsu’s Dedication to the Sustainable Progress Ambitions (SDGs)

The Sustainable Growth Targets (SDGs) adopted by the United Nations in 2015 represent a set of typical targets to be achieved worldwide by 2030.

Fujitsu’s objective — “to make the globe more sustainable by setting up believe in in culture by way of innovation” — is a promise to contribute to the eyesight of a better long term empowered by the SDGs.

The targets most relevant to this job

About Fujitsu

Fujitsu’s function is to make the environment additional sustainable by building trust in society as a result of innovation. As the electronic transformation husband or wife of choice for customers in more than 100 nations around the world, our 124,000 workers perform to solve some of the best issues struggling with humanity. Our array of companies and methods attract on five important technologies: Computing, Networks, AI, Info & Protection, and Converging Technologies, which we convey together to produce sustainability transformation. Fujitsu Restricted (TSE:6702) noted consolidated revenues of 3.6 trillion yen (US$32 billion) for the fiscal year finished March 31, 2022 and stays the prime digital providers company in Japan by industry share. Uncover out far more: www.fujitsu.com.

Push Contacts

Fujitsu Minimal

Community and Trader Relations Division

Inquiries


All business or merchandise names stated herein are emblems or registered logos of their respective entrepreneurs. Info delivered in this press release is accurate at time of publication and is subject to improve with no progress notice.


Day: 8 November, 2022

Town: Tokyo

Corporation: Fujitsu Limited