X

Boundary Layer Technologies is a pioneering company that designs small hydrofoil cargo ships aimed at disrupting the transportation industry. The company's goal is to create a new class of vehicle that could revolutionize global trade. However, creating new technology is complex. A major challenge they faced was creating simulations that would account for free surface – the boundary between air and water – and cavitation effects on wing optimization. They also needed services that could handle the scale of their designs, as they were creating meshes that took hundreds of gigs of RAM that they could not fit on a local desktop.

ARC, an innovative aerospace company that 3D prints metal rocket engines, was facing a bottleneck in their product development pipeline due to the lack of scalability in computing resources and agility in resource diversity. Their local on-premise HPC system with 50-128 cores at 80% utilization was not sufficient to run the large number of simulations required for their product development. This severely limited ARC’s simulation-throughput and design of experiments (DOE), delaying their potential time-to-market. Faced with the urgent demand for more computing resources, ARC had to decide between investing in a static on-premise HPC system or moving to a cloud-enabled HPC system.

Pinnacle Engines, a company developing and commercializing an ultra-efficient engine architecture, needed to run a wide variety of simulations, including the analysis of a four-stroke internal combustion engine using the computational fluid dynamics (CFD) tool, CONVERGE by Convergent Science. The design of an engine with 30-50% fuel efficiency required many simulations, constant iterations, and the compute power to accomplish all this within project timelines. Running tens to hundreds of models simultaneously required extensive compute resources and access to a large pool of simulation software licenses, equating to large up-front costs and personnel investments. Even with a powerful internal processing cluster, the ability to run parallel jobs was significantly limited.

Exponent, a company providing a wide variety of engineering and scientific consulting services, was facing challenges due to the increasing technological complexity of their clients' demands. Their clients, who are building next-generation consumer electronics, energy, healthcare, and automotive products, were increasingly relying on Exponent’s expertise in computational fluid dynamics (CFD) and thermal management. However, the needs of these projects quickly outpaced the computational capacity of Exponent’s internal HPC hardware. The team needed a solution that could be right-sized for each client and simple enough for new engineers joining the team to learn and use quickly.

AGC, a global leader in glass, ceramic, chemicals, and electronics materials manufacturing, faced several challenges related to workflow and how users actively work with simulation tools and data. Sophisticated microstructure geometry designs needed to be modeled from the molecular-level up to the system level. These simulation models are very computationally-intensive and need to be performed continuously and across each step in the design and manufacturing process. Managing complex processes and workflows was further complicated by software updates. AGC uses a diverse suite of applications like LAMMPS, STAR-CCM+®, and COMSOL Multiphysics® to model different types of physics, and it was difficult to keep up with all of the software installation updates and maintenance. As a global enterprise with the highest security standards and requirements, AGC needed the solution to comply with the strictest IT policies, yet also simplify how users work with data and software tools and obtain resources. Other challenges included integrating security from the on-premise facilities to the cloud, minimizing data movement across the wide-area network (WAN) for hybrid cloud workflows, software licensing, accommodating complex, multi-step workflows, administration of policies, methodologies, software, and management tools, and future compatibility to enable breakthrough in areas such as machine learning and deep learning.

Richard Childress Racing (RCR) is a successful NASCAR team that designs and builds its race cars from the ground up. Over the past 12 years, RCR has invested significantly in computational fluid dynamics (CFD) to develop a deeper understanding of the aerodynamics of their cars, evaluate new aerodynamic concepts, and analyze phenomena not modeled in the wind tunnel. However, CFD is a compute-intensive process, and RCR's on-premises resources were limited. They needed to augment their capacity to build larger models with a resolution high enough to precisely understand the intricate flow details that affect the car’s speed on the track.

The global automotive manufacturer was already using multiple HPC cloud services providers to power its digital engineering. However, the company wanted to further streamline its operations by moving its engineering workstations to the cloud. The challenge was to do this without incurring unnecessary costs for idle workstations. Traditional physical workstations were limited in their compute capacity and required significant time to boot up and configure. The company wanted to avoid the costs of cloud compute services when they weren't using the workstations, which was a significant portion of the total time to complete an experiment.

In the automotive business, major automakers send out requests for bids to equipment manufacturers when they need a new part. These requests typically have very short timelines, requiring a response in just a few weeks. The potential business from these contracts—often for millions of parts—is central to the success of auto parts manufacturers. It is essential that they can quickly engineer new equipment designs that perform to specifications, are reliable, and can be sold at a profitable margin. One of Rescale’s customers faced this exact challenge. The R&D teams for this auto equipment manufacturer are at the center of its efforts to develop new designs to win contracts with the major automakers. So it is imperative the company does anything it can to better power its engineering and design process. Without the necessary compute power, the automotive manufacturer was limited in its abilities to fully research and test its new designs before making a bid.

TEN TECH LLC, an ITAR certified company, often services the highly confidential and secure sector of military and government organizations. They require a flexible yet highly secure platform for running models for their military and defense contractors. National and international clients and deadlines mean that they need access to extensive compute resources with the capability of handling large CFD and FEA simulations. A public model TEN TECH LLC recently ran was an analysis studying the effects of hurricane-force winds on large telescope arrays, including stress, deflection, and aeroelasticity. With their local server used for other projects, TEN TECH LLC needed an immediate solution capable of handling their compute-intensive project quickly.

Liberty University's engineering program, one of the fastest growing schools within the university, required computation-intensive R&D and faced challenges providing adequate high performance computing (HPC) resources. The HPC software and hardware required varied greatly across a wide range of engineering programs, including civil, electrical, industrial, mechanical, and computer engineering. Combining that variation with the ongoing need for support posed a challenge for the IT team, so they began to search for a solution that could quickly scale to the needs of the school. They had concerns around cost and their ability to deliver it quickly. When they explored their specific requirements they knew they were short on personnel to implement and maintain a solution on-premises, so they explored options for cloud HPC.

Boom Supersonic, an aerospace startup, is designing a supersonic passenger jet that will revolutionize business travel. However, creating a technological breakthrough in the aerospace sector traditionally requires billions of dollars in R&D, a large engineering staff, extensive wind tunnel testing, and many years of development. Advanced fluid-flow and mechanical stress simulation tools have reduced physical testing requirements and costs, but they require costly dedicated high performance computing (HPC) resources to be effective. As a small startup with limited funding but outsized dreams, Boom turned to the public cloud because the upfront costs of building their own on-premise HPC cluster were cost-prohibitive.

Nissan, a global full-line vehicle manufacturer, was facing challenges with its on-premise High Performance Computing (HPC) systems. The company was limited by fundamental aspects of on-premise computing, such as limited electric power, high total cost, and data center utilization challenges. Their on-premise HPC systems were constrained by the initial hardware and software specifications, were complex to operate, and struggled to handle high-demand (peak) loads. These inherent on-premise problems threatened Nissan's innovation, market leadership, agility, and time-to-market.

DENSO, a leading automotive and Fortune 500 company, operates globally in 35 countries and regions around the world and manufactures a wide variety of components. In its non-automotive business, DENSO is working to industrialize smart agriculture using factory automation and sensing technologies. Within the thermal management systems team, a core businesses at DENSO, new initiatives are underway to respond to electric vehicles and automated driving. Product development for thermal systems in electric vehicles involves the development of a key element of the thermal system, the product development in the thermal systems requires unconventional technologies and methods. It is essential to consider a variety of factors when undertaking new designs. Because of the limitations of manual checking through prototyping, simulation is becoming even more important to improve development efficiency. Mr. Ogawa has championed the development and promotion of simulation methods and tools in the Heat Exchanger R&D Division, which develops heat exchangers used in air conditioners and radiators, key components of thermal systems, and he describes the difficulties of utilizing simulation.

RWDI, a multidisciplinary engineering consulting firm, was facing a challenge due to its business growth. They had outgrown their on-premises High Performance Computing (HPC) environment and needed a solution that could handle their increasing workload. They had recently won a large wind speed mapping project across the Middle East, which required analyzing 30 years of atmospheric data across the entire region in two months. This required a million core-hours, significantly more than their on-premises system capacity. Due to the short project timeline, expanding their on-premises HPC was not feasible. Building out their own software and middleware infrastructure to be able to burst to the public cloud on their own was also not an option. They needed a turnkey hybrid cloud solution to expand their existing capacity.

Engineering world-renowned transportation requires high powered computing resources and access to industry-leading software. When internal capacity is reached, and there are tight project deadlines, Trek turns to Rescale to instantly expand their resources. In a recent analysis, Trek engineers ran a complex simulation using CFD tool Star-CCM+ by CD-adapco and optimizer HEEDS by Red Cedar Technology in a coupled analysis on Rescale’s cloud simulation platform. The goal was to study varying bicycle drafting methods to introduce new angles to the existing analysis method.

Specialized Bicycle Components, a leading manufacturer of performance bicycles, was facing constraints in their internal computing resources which were slowing down their product design cycles. The company's R&D teams were experiencing slow solve speeds which were limiting their ability to innovate and create new product designs. The company operates a proprietary wind tunnel for performance testing, but they also employ computer-aided engineering. However, their existing simulation workstations were becoming insufficient for the growing high performance computing needs of their Road Bike R&D team.

Cambridge Consultants, a world leader in disruptive innovation and technology-based consulting, was facing challenges in streamlining their lab operations due to the high turnover of projects. Their labs needed to be flexible and capable of developing and executing robust new protocols quickly. They required a scalable, robust tool that could provide end-to-end connectivity of devices and data to run protocols effectively, particularly complex multi-factorial experiments with many varying conditions. One area where they needed significant improvement was in the rapid and flexible development of liquid handling automation workflows. They needed a solution that could handle high throughput, fast execution speed, custom sample handling steps, robust sample traceability, excellent reproducibility, and minimized execution costs.

CSL set a goal of doubling the throughput of experimentation via robotics, automation & digital connectivity without a significant increase in headcount by 2030. The CSL Purification Team was looking for a solution that would enable scientists to plan and execute micro-scale purification experiments without the need to write code, ensuring their scientists could increase their throughput while improving the quality of their results. The team faced obstacles such as low uptake of automation and complicated scripting, low device utilization in lab environments, and manual management of experimental workflows & associated data. The team required a tool that would help them standardize their workflow design, execution, and data analysis processes for micro-scale purification experiments and easily transfer knowledge from one site to the next.

InterContinental Hotels Group (IHG) utilizes four different public clouds due to the size and complexity of their IT infrastructure. They wanted the flexibility of multiple public clouds to use the optimal environment for each workload and to protect themselves against price increases and instabilities in any one public cloud. They had been using VMware's vRealize Automation for managing VMware servers, but it did not provide support for the public clouds that they needed. Furthermore, they found vRA to be extremely time-consuming to maintain, with upgrades requiring a multi-month process and a large professional services cost. The process of installing multi-server, multi-tier apps on any one of these public clouds was onerous and required the administrator to use multiple different interfaces.

Before CloudBolt, the enterprise relied on a complicated integration of vCloud Orchestrator, Chef, and a ticketing system to run their server provisioning processes. They struggled with slow turnaround time, a lack of governance, and difficulty understanding their costs. The time required to provision a single VM was reduced from one business day to just 15 minutes. The time to onboard new hosted clients was reduced from seven days to under 30 minutes.

Neustar relied on an internally developed provisioning platform to build resources in a new state of the art data center. However, due to other critical priorities, Neustar administrators were unable to allocate the time needed to extend and maintain the system. As a result, IT began to rely on AWS more and more to meet the service delivery demands of their customers. This resulted in a hefty AWS bill and an expensive datacenter being underutilized. Neustar needed a solution that would better leverage the existing capital investment and deliver services to their customers in a timely and simple manner. Specifically, an “AWSlike” experience combined with the necessary visibility and control of costs and assets.

Before CloudBolt, Blackboard relied on a complicated integration of vCloud Orchestrator, Chef, and a ticketing system to run their server provisioning processes. They struggled with slow turn-around time, a lack of governance, and difficulty understanding their costs. The time required to provision a single VM was reduced from 1 business day to just 15 minutes. The time to onboard new hosted clients was reduced from 7 days to under 30 minutes.

In 2014, Home Depot's IT team realized that they were spending an average of 1-2 weeks to provision VMs for their end users and had a growing shadow IT problem. They decided to evaluate cloud management platforms (including CloudBolt, VMWare’s vRA, Cliqr, CSC Matrix, and a couple of others) to see if one of these CMPs could help them address their challenges. After an exhaustive evaluation, they chose CloudBolt and told us that they were impressed by its simplicity, extensibility, elegance, and its low overhead to install, upgrade, and maintain.

The university’s Office of Information Technology cloud team was tasked with implementing various cloud-based services and exposing them to students and staff in a controlled and reliable manner. They needed a unified interface that would allow end users to easily find and request services. They wanted to enable the use of both public cloud providers and existing university assets while tracking and reporting resource usage to better understand and manage respective costs and capabilities. The team was also faced with the challenge of reducing VM sprawl through active system lifecycle management and repurposing IT resources to advance the university’s educational mission.

The IT group at the Fortune 500 US-based retailer primarily services the e-commerce group, which is responsible for the web storefronts of all the corporate brands, and for a significant portion of the company’s revenue. To speed delivery of new and innovative capabilities, the e-commerce team needed a platform that would allow on-demand access to production and development/test resources, with high compliance to PCI and other standards. Initially, they experimented with a homegrown solution based on OpenStack, but this was rejected by the CIO because it failed to effectively take into account the needs of other lines of business. The potential for infrastructure fragmentation, extra costs, and large amounts of duplicated work caused this company to search for a better solution.

NBCUniversal (NBCU) is an international film and TV studio and distributor. To support the growing consumer shift to digital and mobile consumption of media and entertainment, NBCUniversal knew they needed to have better insight into their software development process and bring developers and operators closer together. They leveraged Splunk to deliver better insights on their internal processes, but soon ran into limitations on their architecture. NBCU’s Splunk application was a critical tool for their DevOps efforts, but their existing environment was underperforming – it was limited in its architecture to ingest 1 TB per day. Like many organizations, their data sets were constantly growing. 1 TB/day wasn’t enough — they wrestled with a huge backlog of unprocessed data and had lost the ability to drive actions in real time. The limits of their physical infrastructure and deployment methods were limiting scale.

As New Relic anticipated their future growth, they saw it as an opportunity to rethink their own application architecture. This led New Relic to morph their existing monolithic application components into a service-oriented architecture with a goal of increasing their number of software deployments each day. At the same time, they were planning to greatly increase their infrastructure and launch new products and features into the marketplace. As a SaaS company, these transitions were not taken lightly, as any changes to the existing infrastructure and services instantly touches their customers. To aid in this transition, the Site Engineering group, including – Sean Kane Lead Site Reliability Engineer Karl Matthias and Sean Kane, along with several product teams, started to investigate Docker, the open platform to build, ship and run distributed applications.

Sleepy Giant, an independent online game company, was in need of a public cloud solution that could provide flexibility in scaling infrastructure up or down at a moment's notice. The gaming industry is characterized by unpredictable usage spikes due to organic, viral uptake and paid promotions. When these spikes occur, it is crucial that infrastructure be quickly scaled up. If increased usage is not met with adequate infrastructure, then players suffer from decreased response times, the game experience deteriorates and opportunities are missed. Sleepy Giant also required a support team with broad knowledge and competence to help build the right infrastructure for its partners. They were seeking a team that could provide customized solutions and recommend proven options to meet the continuously expanding needs of its diverse projects. Lastly, Sleepy Giant wanted to reduce infrastructure costs while continuing to deliver improvements in performance to customers. This meant considerations of alternatives to ownership and management of physical hardware.

Wanelo, a community for all of the world’s shopping, was experiencing a surge in popularity and traffic to its website and mobile apps. The company wanted a cloud environment that was designed for performance and could cost-effectively handle traffic surges. The company's growth, expressed in terms of requests per minute (RPMs), took off in a period of approximately four months following a major site rewrite. Wanelo wanted a cloud environment that could handle the evening surges without a cost penalty or needing to overprovision the system to accommodate those times of peak traffic. They also wanted more OS functionality than Linux.

Digital Chocolate, a leading provider of social games, was facing a challenge with its growing daily active users (DAUs) base. As the DAU base grew, so did the company's infrastructure costs. Initially, supporting the explosive DAU growth was the company's primary focus. However, as the business evolved and the complexity of supporting games increased, it became clear that Digital Chocolate required a more flexible solution to accommodate its needs. The company needed to ensure that its costs per DAU didn't exceed its revenue per DAU. With Digital Chocolate's incumbent cloud infrastructure provider, sudden bursts of bandwidth usage often drove up costs, jeopardizing the profitability of some games. The company also needed to ensure maximum performance and adaptability to its needs.