Dinex er en dansk designer og fabrikant af udstødnings- og emissionssystemer til lastbiler, busser og off-road-køretøjer med produktions og salgsfaciliteter i 16 lande over fire kontinenter. Deres produktionsproces er overvejende baseret på simulation af udstødningskomponenter og subsystemer, CFD-simuleringer af konjugeret varmeoverførsel fra væske til struktur; FEA-simuleringer af termiske belastninger på termiske dele, og FEA-simuleringer af strukturelle statiske og dynamiske kræfter. De kører CFD på AVL, FIRE og FEA på ANSYS Mechanical. Med 4 – 5 millioner celler pr. model er disse modeller store, og kører typisk i tre uger på deres 32-kerners in-house-server.

Dinex har haft et projekt for en kinesisk lastbilproducent i kommission for at finde frem til et nyt selektivt katalyseudstødningsreduktionssystem (SCR), med hvilket producentens lastbiler kunne opfylde Kina 5-emissionsnormerne.

Ved projektet blev der krævet en meget stor model, men også en kort tidsfrist. Med kun et par uger til at gennemføre projektet anvendte Dinex skyen for at gøre brug af computer-ressourcer.

Rescale-løsningen

Dinex overvejede både at investere i en større in-house-server og at benytte Microsofts Azure i skyen.

De besluttede sig til ikke at anvende en in-house-server, fordi det ville have taget for lang tid at opbygge en sådan, den ville have været for dyr og ville måske ikke have kunnet klare deres fremtidige krav.

Azure gav fleksibilitet i skyen, men var for kompliceret at implementere. I modsætning til de ovennævnte muligheder udgjorde Rescale en nøglefærdig løsning for både bestilt hardware og software til betaling pr. anvendelse, og havde et letanvendeligt arbejds-flow. Summa summarum valgte Dinex Rescales ScaleX-platform på grund af dens enkle funktion og skalerbarhed.

Til projektet kørte Dinex seks design-optimeringer af en multi-faset urinstofsimulering for at forbedre ammoniakensartetheden og reducere risikoen for aflejringer vha. AVL FIRE. Hver optimering omfattede to simulationer i parallel. Med fire til fem millioner celler kørte simuleringerne på ca. 100 kerner hver. Under anvendelse af Rescales store beregningsplatform til at skalere ud horisontalt tager hver optimering nu ca. tre dage at køre, i sammenligning med tre uger på deres in-house-kontrolelement.

“Vores højere produktivitet vil sætte os i stand til at håndtere flere projekter. Rescal-platformen gør os mere effektive og mere attraktive for vores kunder, hvilket igen giver os mere omsætning og flere udviklingsprojekter.”

Kasper Steen Andersen CAE-chef, Dinex

Resultater og fordele

Ved at benytte Rescales ScaleX-platform kunne Dinex reducere leveringstiden for hver optimering fra fire uger til en uge, svarende til en reduktion på 75 % i produktionstid. (Hver iteration omfatter ændringer i design, oprettelse af nettet, opsætning af modellen på ScaleX, kørslen af den, gennemsyn af resultaterne og beslutningstagning). I alt reducerede dette simulationstiden for hele projektet med 18 uger.

Hurtigere simulation gav mange væsentlige fordele for Dinex, inklusive:

• 25 % reduktion af tiden til at et udstødningssystem kan komme på markedet. Med Rescale-platformen blev den total projekttid reduceret fra 18 måneder til 13,5 måneder. Dette gjorde at Dinex kunne overholde en kort leveringsfrist og overgå klientens forventninger.

• Højere omsætning af projekter og mere indtjening i kraft af deres kapacitet til at påtage sig flere projekter.

• Mere tilgang af produktivt personale. Simulationsingeniørerne begyndte at arbejde mere effektivt og forøge deres produktion, nu hvor simulationsflaskehalsen var overvundet. Dilnex’s ledelse påregner at ansætte flere simulationsingeniører for at kunne holde trit med den hurtigere simulationsomsætning.

• Bedre markedsføringsmuligheder over for kunderne. Ved at anvende Rescales løsning i skyen beviser Dinex at de er på forkant med simulations-trenden ved at benytte branchens nyeste teknologier til frembringelse af optimerede designs på rekordtid.

Om Dinex A/S

Dinex er grundlagt i 1982. Til dato har selskabet 1.250 ansatte over hele verden i mere end 16 lande, med hovedkvarter i Danmark, 8 fabrikker og 7 salgs- og distributionsselskaber. Som leverandør af komplette systemer er Dinex som producent af efter-markedsudstyr (AEM) og producent af originalt udstyr (OEM) den markedsførende i Europa. Selskabet har en lang historie koncentreret om det globale heavy-duty-marked med en teknologisk og globalt produktiv platform.

18% Cost Optimization of Applications and Productivity

•  With the latest computing resources on-demand, Nissan was able to match utilization effectively and optimize the simulation performance for specific applications
• The flexibility to capture every possible opportunity for cost savings, utilization, and allocation 

50% Cost Reduction for HPC Expenses 

• With the pay-per-use model, Nissan could minimize the HPC cost of on-premise systems during peak usage when including the cost of electricity, operations and Human Resources
• They avoided over-provisioning and wasting resources of fixed capacity on-premise systems 

0 Queue During Peak Demand

• With improved infrastructure agility – Nissan’s capability to quickly deliver the right computing resources, in the right scale, and precisely when it is needed without queuing or wasting engineering labor
• With on-demand cloud resources, Nissan reduced queue time from 2-3 days to zero days

About Nissan

85 Years of Innovation and Leadership

Nissan is a global full-line vehicle manufacturer that sells more than 60 models under the Nissan, INFINITI and Datsun brands. In fiscal year 2018, the company sold 5.52 million vehicles globally, generating revenue of 11.6 trillion yen. Nissan’s global headquarters in Yokohama, Japan, manages operations in six regions: Asia & Oceania; Africa, the Middle East & India; China; Europe; Latin America; and North America.

“The challenges we faced at Nissan revolved around managing our in-house HPC system alongside keeping up with the latest technology innovations needed to meet market demands” – Seiji Kawachiya, General Manager of Engineering and Quality System Department, Nissan Motor Corporation.

Shifting to an Agile Innovation Model

Nissan chose to shift its technical computing to Rescale to gain an agile, cloud-enabled, platform-based solution on a pay-per-use model to minimize overall cost-per-simulation, gain instantaneous agility, and continuously adopt the latest technology. This allowed Nissan to:

• Allocate the right resources at the right time
• Avoid over-provisioning and wasting resources
• Avoid under-provisioning and creating product development delays

Challenges with On-premise Computing

Like many enterprises, Nissan realized they were limited by fundamental aspects of on-premise computing, such as: limited electric power, high total cost, and data center utilization challenges (Nissan averaged 40% – 80% utilization). Nissan recognized that these inherent on-premise problems threatened its innovation, market leadership, agility, and time-to-market.

Their on-premise HPC systems also faced the following challenges:

• They were constrained by the initial hardware and software specifications. Nissan’s computing investments were fixed over a 4 year cycle, until a 9 month hardware refresh.
• They were complex to operate. Nissan’s HPC systems required many specialized employees, facilities, application maintenance, and security measures.
• They struggled to handle high-demand (peak) loads. Nissan’s fixed HPC resources caused their engineers difficulty in allocating priority workloads to computing resources, and added risk of missing design improvement opportunities.

“The alternative options that were available outside of Rescale such as building our own solution or extending our on-premise environment were inadequate and did not meet the full demand of what Nissan required.” – Dai Matsubara, Assistant Manager of Engineering and Quality System Department. at Nissan, System Director of Nissan HPC.

Committing to Rescale

To address these challenges, Nissan chose Rescale’s ScaleX Enterprise platform because it provides diverse and scalable HPC resources on a pay-per-use model, and an efficient global administration infrastructure. Nissan also valued that the Rescale solution provided:

• The ability to instantly scale a diverse range of the latest hardware and applications
• Controls for administering simulation users, projects, software, hardware, and budgets
• A single solution that supports Nissan’s journey to the cloud, starting with hybrid

The Rescale Solution

Rescale’s managed HPC platform converted Nissan’s complex stack of software, hardware, and infrastructure into a single, unified solution that remains agile in the fast-moving technology environment. Nissan’s engineers are able to access the latest technology selection through a single, unified solution and at virtually unlimited scale.

Agile Engineering and IT Business Impacts

Within 4 months, Nissan’s global engineering division was concurrently running hundreds of jobs per day on the Rescale platform. They experienced the following results:

• Completely eliminated all queuing time, which was previously an average of 2 – 3 days
• An 18% cost optimization of applications and productivity due to better matching the hardware to specific workflows. Nissan expects a further cost optimization of 15% with next generation hardware, bringing the combined cost and productivity improvement to more than 30%
• 50% cost reduction on HPC operating expenses

With 3 – 4 times higher software cost than hardware, shifting to better optimized hardware and agile scalability allowed Nissan to achieve more with the same licenses and improve their workflow efficiency.

Assessing the Complete Cost of HPC in the Cloud

Nissan carried out in-depth cost assessments when considering alternatives and was able to combine many different views (IT costs, engineering costs, products, and safety risks). With the latest hardware and unlimited computing resources on demand capacity, they reduced time to market for their newest innovations.

Nissan leased on-premise HPC systems at a fixed rate over the life of the hardware. This made on-premise hardware appear less expensive than cloud for Year 1, however the overall cost of an agile cloud platform solution became substantially lower.

Conclusion 

Nissan recognized that its ability to remain agile and innovate faster was their key competitive differentiator in the automotive markets. They saw it was a strategic imperative to deliver their global engineering team with platform HPC resources to increase overall operational efficiency. Nissan chose the Rescale ScaleX Enterprise platform, which successfully delivered improved ROI across these business drivers:

•  An 18% cost optimization of applications and productivity, achieved by using optimized resources for each application and the agility to instantly scale resources to match computing demands.
• One unifying platform solution for long-term sustainable product quality and innovation enabling rapid new technology adoption.
• Faster product development, achieved through improved infrastructure agility and turnkey global deployment and administration, together provide zero queue time during peak engineering demand. 

For more detailed information, please contact [email=sales%40rescale.com]sales@rescale.com[/email].

The Rescale HPC infrastructure management platform includes over 700 of the most widely used and powerful engineering and scientific applications. Furthermore, Rescale partners with the four largest cloud service providers.

With the speed of HPC software and hardware development, accessing the best package for individual workloads is practically impossible, negating one of the many reasons to switch to cloud. Rescale is the cloud HPC orchestration platform, making price performance optimization at a job level possible. Rescale offers hundreds of software applications and dozens of cloud infrastructure architectures available for engineers to develop models and run simulations in several engineering disciplines.

Rescale offers powerful collaboration tools for teams distributed around the world. Administrators and project leads can quickly provision specific access to resources and data within shared workspaces to ensure downstream tasks can be carried out without delay. With Rescale, businesses can compose and automate common simulation workflows, reducing costly down-time from new product development. By simplifying HPC operations, engineers can focus on their core competencies and not worry about obtaining, setting up and optimizing their computer software and hardware.

https://rescale-hpc-cloud.com/

“Computational Science & Engineering is giving early adopters competitive advantages in every major market around the world today,” said Edward Hsu, Chief Product Officer, Rescale. “But most workloads remain on premises. Where we see the early and most promising design and discovery breakthroughs is where organizations unlock computational barriers and embrace the cloud. Rescale provides the leading platform that enables those breakthroughs while providing IT the security and control they need.”

Computational Science & Engineering includes several important computing industries where companies and governments need computational models to simulate and understand natural phenomena such as weather and quantum mechanics or how engineered products will perform under enormous stresses such as aerodynamics and crash tests. This domain includes standard computer engineering, computer science, high performance computing and supercomputing. Rescale’s platform also allows its customer to run workloads on the world’s fastest supercomputer, Fugaku by RIKEN in Japan, as well as leading public cloud providers such as AWS, Microsoft Azure, Google Cloud Platform, Oracle Cloud and more.

The new report, available free for download here, follows Rescale’s publication last year of the “Big Compute 2021 State of Cloud HPC Report” that reported industry analysts forecasting a $55 billion annual HPC market by 2024. Rescale believes the computational engineering and science market may double in size by that time as more and more companies take advantage of its on-premises and hybrid cloud platform to run new workloads in artificial intelligence and machine learning by giving new market entrants easy access to the world’s most powerful computing systems as an operating expense.

The 2022 report reveals that a new era of computational engineering is fostering a second wave of massive disruption and innovation in engineered products across all industries. In driving successful engineering innovations, access to massive computing power, a library of thousands of algorithms and proprietary software applications, and automated workflows with granular insights into operating costs, Rescale aims to remove the last remaining computing bottlenecks for companies of all kinds. By giving engineers easy and unlimited access to computing power that offers up to an order of magnitude of higher performance, companies will be able to explore a much wider range of design possibilities and accelerate industry innovations.

Other key findings in the key findings include:

• R&D leaders know that researchers spend much of their time on non-research related tasks (e.g., finding lost files, setting up infrastructures). But the impact of this non-R&D time on project success is not well understood;
• Organizations that can help researchers focus their time on R&D are more than twice as likely to achieve project goals consistently;
• Easy access to compute accelerates innovation velocity and also improves the probability of project success;
• Ease of access to computing is highly correlated with the ability of organization to tackle broader science and engineering challenges;
• Organizations that use cloud automation platforms generally do so as part of a digital R&D strategy.

Rescale, the leader in high performance computing built for the cloud to accelerate engineering innovation, has been named to Y Combinator’s Top Companies list for the fourth year in a row. 

“We’re honored to be recognized by Y Combinator on their Top Companies list,” said Joris Poort, founder and CEO of Rescale. “At Rescale, we are pioneering a transformation in computational engineering. Our company has helped private startups do things never-before imaginable, like rockets sending satellites and people into space, building supersonic jets and electric vertical take-off vehicles combining for billions of dollars of commercial orders. Our customers are completing these projects in years, not the decades it took before.” 

Founded in 2011 and backed by investors like Jeff Bezos, Sir Richard Brandson, Microsoft, Samsung and NVIDIA, Rescale is accelerating engineering breakthroughs and making it possible for scientists and engineers to do more, faster.  Rescale automates the entire high performance computing stack, empowering users with computational science and engineering applications with the world’s most powerful computing architectures, while giving IT security and control.

Listed and ranked by valuation, Rescale joins other Y Combinator alumni like Airbnb, DoorDash, Reddit and Stripe on the Top Companies list. This year’s list includes 267 public and private companies from 20 countries, most valued at more than $150 million, with more than 60 valued at more than $1 billion. An interesting trend this year shows more than 11% of the companies listed are remote first, with Rescale also looking to fill more than 100 positions in 2022 globally, many of them remote.

Y Combinator is the largest startup accelerator in the world, investing in a large number of startups every year, and then working with those companies to attract investor interest from the world’s leading venture capital firms.  Rescale completed the program in 2012.

The Top Companies list selects from the more than 2,000 companies they’ve worked with since 2005.  This year marks the fourth time Y Combinator has published the Top Companies list, with Rescale being listed every year.

Containers are popular for delivering workload portability and efficiency, but high-performance computing use cases bring additional requirements on security and parallel processing. Rescale has long supported the use of containers to bundle applications and their dependencies needed to execute on the best possible architecture and run those workloads with 800+ applications on Rescale’s always-current software catalog. 

Containers simplify the development and deployment of custom software because they abstract the complexity of hybrid and multi-cloud deployments and dependencies that can arise from differences in infrastructure environments. Rescale integrates the leading container technologies to enable open-source collaboration and proprietary applications development. Growth in home-grown application development, for example proprietary AI/ML models, has spurred further adoption of containers on Rescale to achieve increased engineering productivity and enhanced IT control.

The benefits of using Rescale to run containers include:

• Multi-Cloud – Run your containerized applications on any cloud in any region.
• HPC-Optimized – Support more secure and parallel workloads while optimizing for best performance or cost efficiency.
• Integrated Experience – Integrate your container workloads alongside 800+ commercial & open-source science and engineering applications. 

“Organizations powering computational science and engineering with HPC can face daunting challenges of complexity,” said Adam McKenzie, CTO and Founder at Rescale. “Containers on Rescale help to extend the simplicity and full-stack cloud automation we bring our customers into containerized workloads. The rapid growth of AI/ML, open-source, and custom applications is driving a convergence with HPC. Rescale provides a bridge for these technologies to not just run applications but supercharge them with the latest architectures whether you care about optimizing cost or speed while at the same time giving IT the security and control, they require.”

Customers deploy their containerized workloads on Rescale in the cloud to access greater scale and better performance on alternate infrastructure. Containers also make it easier for IT operations to better support and manage their in-house applications seamlessly with their other commercial applications. From autonomous vehicle testing to industrial manufacturing optimization, Rescale customers are accelerating the training of AI/ML models and other proprietary applications using the Rescale platforms. 

In-house HPC software development is often bespoke and run with large data sets across on-premises data centers or specific architectures. Containers make these same workloads portable, which is a big advantage when customers want improvements in performance and price. Containers deliver applications as microservices. These microservices maintain their lifecycle along with service-specific requirements of independent development, granular scaling, and patching and fault remediation. 

Similarly, containers give HPC and AI/ML workloads the advantage of isolated management that includes scaling and development. Container scaling is a major advantage in HPC where workloads can spike in data processing requirements.

Rescale advances in container capabilities include:

• Multi-node high performance support for Singularity and Apptainer containers; Docker container support for single node applications;
• ARM architecture support for the latest Singularity versions (3.9.6);
• High-performance networking, such as InfiniBand using Open MPI;
• Container publishing via Rescale Templates for collaboration and standardization.
• NVIDIA GPU-optimized containers for seamless integration of world-class AI applications.

Rescale customer AGC, a global leader in glass, ceramic, chemicals, and electronics materials manufacturing, founded in 1907, uses containers to support ML and Deep Learning (DL) capabilities using their cloud provider of choice. Containers give the company the portability and compatibility they need to run in-house applications anywhere. Read more about AGC here.

Backed by investors like Jeff Bezos, Sir Richard Branson, Microsoft, Samsung and NVIDIA, Rescale is accelerating engineering breakthroughs in a wide range of industries, making it possible for scientists and engineers to do more, faster. 

“Today, heavy-duty transportation pumps a tenth of the world’s greenhouse emissions into the atmosphere every year,” said Seonghoon Woo, CEO of Amogy. “By adopting the advanced computing techniques of Rescale, we are able to move much faster in our product development cycles. We are seeing a 5X increase in engineering productivity through increased fidelity in our simulations and modeling, while at the same time reducing our compute infrastructure spending.”

Founded in 2020 by a team of former MIT PhDs and based in Brooklyn, New York, Amogy’s goal is to create an ammonia-based power system that would dramatically reduce greenhouse emissions globally – by as much as 10% over the next two decades, Woo said. Amogy’s solution combines advancements in hydrogen fuel cells with the advantages of ammonia as a fuel source.

In December 2021, Amazon’s Climate Pledge Fund invested an undisclosed amount in Amogy. Partnering with Amazon, Amogy will work towards the demonstration of an ammonia-powered clean cargo-shipping vessel. Amazon has said that decarbonizing maritime shipping is one of the biggest environmental challenges that has to be addressed to tackle climate change.

Amogy’s technology development relies on computational research using high performance computing. Engineers at Amogy test multiple models, with a variety of parameters and under varying conditions. A key goal is to optimize a design that maximizes the ammonia-to-hydrogen conversion process in their reactors. The ultimate solution – an alternative energy replacement for fossil fuels that power heavy transport – required computational resources for simulation and modeling at massive scale. Rescale offered a complete HPC-as-a-Service solution with a UI designed by and for scientists.

Backed by investors like Jeff Bezos, Sir Richard Branson, Microsoft, Samsung and NVIDIA, Rescale is accelerating engineering breakthroughs in a wide range of industries, making it possible for scientists and engineers to do more, faster. Rescale automates the entire high performance computing stack, empowering users with computational science and engineering applications with the world’s most powerful computing architectures, while giving IT security and control.

“Our customers and partners use Rescale to dramatically speed up product development cycles to create breakthrough solutions in almost every industry,” said Garrett VanLee, Senior Product Marketing Manager at Rescale. “We’re excited to work with the engineers, scientists, and researchers building clean technologies that transform industry and protect the environment for future generations.”

About Amogy

Amogy offers ammonia-based, emission-free, high energy-density power solutions to decarbonize transportation for a sustainable future. Founded by four MIT PhD alumni with a shared vision in 2020, Amogy aims to enable clean transportation in all sectors, accelerating the global journey towards Net Zero 2050.

Wenn Sie einer Person, die Ihre Website schon einmal besucht hat, empfohlene Produkte zeigen, lassen Sie sie damit wissen, dass Sie sie erkennen und als wertvollen Besucher behandeln. Die Möglichkeit dazu gibt Ihnen ein sogenanntes „Empfehlungssystem“, welches Daten aus früheren Benutzerinteraktionen verwendet. Empfehlungssysteme haben in den letzten Jahren sehr an Bedeutung gewonnen und sind für den E-Commerz eine riesige Chance, insbesondere in Bezug auf Big Data.

HPC-Cloud bietet viele Vorteile On-Demand
Mit der Nutzung großer Datenmengen und den Einsatz ausgeklügelter Algorithmen wird das Ziel verfolgt, die Konversionsrate des E-Commerce durch den Einsatz moderner Empfehlungssysteme zu steigern. Die HPC-Cloud bietet On-Demand-Zugriff über verschiedene Browser auf eine breite Palette von Computerressourcen, darunter Softwareanwendungen, physische und virtuelle Server, Datenprozessoren und Speichergeräte, Software- und Hardware-Entwicklungstools und Datenbanken. Dies bietet mehrere Vorteile, darunter erhöhte Simulationsleistung, reduzierte IT-Kosten, erhöhte Ressourcenverfügbarkeit und Flexibilität.

Benutzer müssen beispielsweise nicht in kostspielige homogene Cluster investieren, die lange Bereitstellungszyklen erfordern, welche sich nicht an Spitzenlasten anpassen. Cloud Computing kann bei Bedarf ermöglichen, dass diese Simulations- und Modellierungswerkzeuge fast sofort zur Verfügung stehen. Die intelligente Empfehlungs-Engine von Rescale, die auf Milliarden von Rechenstunden eingestellt ist, wird für Unternehmenskunden immer wichtiger und wertvoller, da die Auswahlmöglichkeiten für Rechenarchitekturen exponentiell wachsen, sich die Cloud-Wirtschaft schnell entwickelt und täglich neue Anwendungsfälle von Kunden angenommen werden.

Wichtige Ressource für Zukunft des Enterprise-Computing
Die wachsende Komplexität von Algorithmen wie KI und maschinellem Lernen, kombiniert mit der raschen Einführung von Computational Science und Engineering in allen wichtigen Branchen, macht diese Plattform zu einer immer wichtigeren Ressource für die Zukunft des Enterprise Computing. Rescale treibt als führender Anbieter von intelligentem Computing für die digitale Forschung und Entwicklung mit Hochleistungs-Computing für die Cloud industrielle und wissenschaftliche Durchbrüche mit umfassender intelligenter Automatisierung voran, die auf den unbegrenzten Rechenkapazitäten der Hybrid-Cloud basiert.

Die meisten E-Commerce-Unternehmen verwenden eine Cloud-basierte Lösung. Warum? Diese ist sehr einfach. Die Cloud bietet zahlreiche Vorteile, die sich perfekt für geschäftliche Anforderungen eignen. Beispielsweise kann die Cloud Traffic-Spitzen bewältigen und hat eine hohe Servicezuverlässigkeit, während sie einen hohen Datenschutz und andere Vorteile bietet. Darüber hinaus kann die Cloud-Infrastruktur für die verwendete E-Commerce-Plattform optimiert werden, wodurch eine Leistung sichergestellt wird, die herkömmliche Dienste nicht erreichen können.

Kundenakzeptanz und Marktchancen für intelligente Hybrid-Cloud-Plattformen wachsen
Steigende Investitionen in intelligente Hybrid-Cloud-Plattformen fördern einen zunehmend überzeugenden datengesteuerten Netzwerkeffekt, da die Kundenakzeptanz und die Marktchancen in den nächsten zehn Jahren schnell zunehmen. Indem Sie die Metadaten von Millionen einzigartiger Workloads, Zehntausenden von Anwendungen und Hunderten von verschiedenen Rechenarchitekturen nutzen, tragen Sie dazu bei, die besten Empfehlungen für die Bereitstellung der besten Leistung für Ihre Kunden sicherzustellen.