What you need to know
Advancements in genomics are opening new doors for understanding human diseases and are increasingly informing innovative precision treatment plans. However, genomics sequencing generates hundreds of petabytes of data per year. Researchers and scientists require tools to analyze these enormous volumes of data in a timely manner to gain insights into disease and possible treatments. GARAOTUS provide the AI-HPC technologies to accelerate genomics workflows. Performance gains include a 75 percent speedup for the BWA performance.
Mind Commerce, an information and communication technology market survey company, pointed out that the total global market value of high performance computing (HPC) will hit 4.93 billion US dollars by 2025, with the CAGR growing by 8.8% from 2020 to 2025. Having said that, the COVID-19 pandemic starting from the end of 2019 has raged the world and had a direct impact on industries’ capital investment. There are still some exceptions - medicine, AI-related industries, and national research institutes managed to sail against the wind and allocate bigger budgets.
GARAOTUS is a brand like this. When numerous industries in the world are at a disadvantage, GARAOTUS makes the most of its solid infrastructure technologies in AI cloud service and HPC. They can provide industries with distributed computing that is highly integrated as well as centralized management, continuing to introduce services to large cross-national enterprises. This is why they can still stand firm in the post-pandemic era.
Counting on leading technology, GARAOTUS has expanded its territory to the global market ahead of others. Thanks to their preemptive action years ago, more international enterprises have sought to partner with them in the post-pandemic era to gain the benefits brought by AI and HPC.
Key HPC technology that wins favor from international partners
AI, HPC or cloud computing are not new to industries, but why can they keep winning favor from world renowned corporations?
Henry Chang, the senior architect of GARAOTUS, mentioned, “The whole structure of GARAOTUS is divided into three key techniques: AI computing, HPC service, and cloud service. These three building blocks can not only provide independent service, but can also be integrated into the GARAOTUS Cloud Platform to generate a centralized management platform for customers.”
We have been aware that one of the features that AI Computing can deliver is a server equipped with graphics cards. The mainstream of GPU technology is led by Nvidia who has realized the environments of deep learning and machine learning through its software packages and programming language; however, GARAOTUS can distinguish the scenarios of AI Computing from those of traditional HPC. For example, previous applications of NLP and the deep learning of images can be solved by adopting the AI computing structure.
In early times, the HPC structure was a parallel computing environment made up purely of stacked multiple physical machines. On account of the technology development and constant breakthrough of x86 and CPU core count, the novel HPC structure thinks outside the box and blends cloud and container technology as the foundation of high-flexibility HPC. This approach has been infused with more possibilities of HPC as a Service.
Jeff Chang, the director of technology consultant, stated, “When global players fell over each other to build cloud technology back from Infrastructure as a Service, to Platform as a Service, and to Software as a Service back in 2014, Jay Hsueh, Head of GARAOTUS, sniffed the trend of cloud development. He then pushed forward the launch of the first generation of GARAOTUS Cloud Platform which offered housing colocation.” GARAOTUS has now made further progress and created Cloud Orchestration that enables the next generation of GARAOTUS Cloud Platform to produce the highest performance from different cloud services.
GARAOTUS also conceived the idea of Cloud Orchestration Flow based on past practices, along with the future trends of technology and application. This can satisfy the service procedure required for relevant industries around the world. From embarking on data collection in the first place or existing BigData datasets from customers or IoT, they are the one with solutions to the system architecture at the bottom for AI data analysis modelling. Furthermore, the HPC environment is of use to stimulate distinct scenarios and applications and further runs storage and simple application algorithms with IaaS in the cloud. The operators will gain access to the information system deployed and applied automatically. In addition, the customers will receive a suggestion about a calculated model to be used for servitization so that the stages which every component goes through can be X as a Service. That is, the goal is Everything as a Service.
Striking performance and successful case examples with international partners
Under the circumstances of the pandemic, the medical industry estimates that there will be massive demands for AI application in the future. According to the report of a survey and research body, the market size of AI medicine reached 2.2 billion US dollars in 2018, and it is projected that the figure will amount to 17 billion US dollars by 2024. AI has already been applied in precision medicine, preventive medicine, and clinical trials all over the world, and meanwhile, the medical AI race has been sparked among countries. Many medical institutions in Taiwan are also proactive in the deployment as big data and cloud technology will lead the healthcare industry to digitalization and intellectualization. The medical system in Taiwan ranks among the best, and continues to accumulate a large amount of medical data. How they can leverage these data, realize more groundbreaking practices, and expedite the development of medical application are the challenges facing the medical world in Taiwan.
Marissa Lin, the senior architect of GARAOTUS, mentioned, to targeting the fast-growing genomics market, GARAOTUS partners with Intel and Advantech, to deliver integrated hardware solution and deployment service. GARAOTUS partners up with Intel for the Genomics Analytics solution, making use of Intel technologies and Broad Institute’s Genome Analysis Toolkit (GATK). GARAOTUS provide the AI-HPC technologies to to accelerate genomics workflows. Performance gains include a 75 percent speedup for the BWA performance.
Advancements in genomics are opening new doors for understanding human diseases and are increasingly informing innovative precision treatment plans. However, genomics sequencing generates hundreds of petabytes of data per year. Researchers and scientists require tools to analyze these enormous volumes of data in a timely manner to gain insights into disease and possible treatments.
Director Jeff Chang also shared a real case example in the medicine field, indicating that “the medical research teams of Taiwan show a high level of competence, but, in the past, before they published new findings, other countries beat them to it. This happened a lot primarily because of the insufficiency of internal computing resources, so they are not able to produce results right away. Besides, the research resources that each laboratory has are separated from each other, which results in ineffective exchange and utilization. In light of this, the customer decided to invest substantial funding, aiming to become a world-class AI medical research and development center.” Director Chang added, “We built the GARAOTUS Cloud Platform for the customer to integrate all of the research resources in the medical system and provide plentiful laboratories with a centralized working environment for computing. In such a manner, researchers can focus on high intensity computing and artificial intelligence applications in biomedicine, supporting the study requirements for image analysis and big data analysis. This is to pursue the optimal solution where AI big data analysis and computing capability come into play to the largest extent. With regard to the architecture, we do not merely take future computing and storage expansion into consideration, but also develop a geo-redundant and data backup system dedicated to the protection of project data, with a cleverly comprehensive design for connection and information security.”
This enormous and benchmark medical investment project is believed to have attracted world leading information giants to take part in the bid. In this fierce competition, only few businesses entered the stage of Proof of Concept. However, the solutions proposed by many world leading corporations were mostly built-to-suit products or customized systems that were assembled temporarily. By contrast, the solution from the GARAOTUS team comes with a mature platform that offers service tailored to customers’ needs while thoroughly inspecting customers’ overall planning of architecture and future development and application. This is also the reason that many international enterprises compete for collaboration with GARAOTUS.
Speaking of the POC process, Director Jeff Chang could not help but raise his voice, saying, “Given that the resources of each laboratory were scattered, our team had to visit each one for an interview in order to coordinate three POC topics identified by the customer into application scenarios. At that time, we made a major breakthrough after operating a certain algorithm on a GPU server. It had required 200 hours to run in a CPU server environment, but it took no more than 40 minutes with a GPU server. The efficiency rose 300-fold; accordingly, the customer was extremely satisfied with the result.”
Three compelling reasons for enterprises to take early action and “move your cloud forward” with GARAOTUS
Though GARAOTUS is a brand new cloud platform technical team, its vision of service is Moving Your Cloud Forward which explicitly underpins its determination to continuously develop and deepen the technology. Moreover, the four words, Moving Your Cloud Forward, also fully explains its sense of mission of lending a hand to partners at home and abroad and marching forward together in the most challenging times.
GARAOTUS has cultivated the market of diversified platform applications, namely virtualization, cloud computing, big data, AI, container, edge computing, cloud service, and cloud native for a couple of years. Their key advantage of developing GARAOTUS Cloud Platform is also the reason that urges large international enterprises to choose the brand.
First of all, the platform is capable of a record level of system integration. If a business is confronted with technical requirements for integrating AI computing, HPC service, and cloud service, turning to GARAOTUS’ product and service will be the most direct and effective solution.
Secondly, GARAOTUS’ product and service is the optimal option for customers to have a servitized information system at hand. The software and hardware under control will display the best performance and benefits through the system, which significantly minimizes both time and cost for building for customers.
Last but not least, information or technical personnel are likely to get the hang of it quickly, so it takes less time to learn. Therefore, the platform has separated the wheat from the chaff, to let general users be able to pick it up easily.
Cloud application service has rapidly spread across the globe, facilitating the industry’s evolution of new technology from the old. The GARAOTUS team has set the pace in AI cloud and HPC area with their leading technology. No matter what plans are carried out to combat diseases or develop medicine, this technology will arm a variety of research institutes with more advanced weapons, and considerably improve the competitiveness of the medical academia and medical industry. Stepping forward, Taiwan’s AI medicine will be empowered to shine on the global stage as the pride of Taiwan.
This article is provided by SYSTEX and edited and approved by The News Lens Brand Studio.
Learn more about GARAOTUS：