▲Taiwan AIoT launched a one-stop AFS POD computing power deployment service for enterprises, helping them significantly reduce the time and cost of building and maintaining AI data centers, and promoting the implementation of AI innovation.
Generative AI, which is making rapid progress in various fields, has become a driving force for industrial upgrading and transformation. It is attracting more and more companies to integrate this technology into their core competitiveness of the new 4Ps, in order to empower employees and customers (People), optimize operational processes (Process), innovate products and services (Product), and thus regain their corporate position in the market and achieve a leapfrog development.
While many public cloud providers offer generative AI services, industries subject to strict regulations, such as finance and healthcare, or high-tech industries with a large amount of confidential research and development, are increasingly choosing to build their own AI data centers after considering factors such as "data privacy and security," "simplified operational processes," "product and service optimization," and "long-term cost-effectiveness." This involves deploying large language models within the company to ensure critical privacy protection, effectively reshaping the enterprise's new 4Ps, and building competitive barriers.
Taizhi Cloud has observed that enterprises remain concerned about uploading data to cloud platforms for training and inference, fearing that the data might become part of the training process, leading to the leakage of important trade secrets. Therefore, more and more companies are choosing to build their own AI data centers to balance the protection of confidential data with the needs of AI projects. Taizhi Cloud's AFS POD service can significantly shorten the time required to build an AI data center and utilize various AI management tools to accurately use AI resources, promoting the implementation of innovation.
▲AFS POD service provides enterprises with a one-stop solution for the GPU computing power, models, applications and processes needed to implement AI, and can be launched in as little as 8 weeks.
One-stop complete service to create the best solution for AI data centers
Driven by factors such as compliance with regulations, protection of confidential data, and enhancement of corporate competitiveness, enterprises have a strong desire to build their own AI data centers. However, they also face challenges such as talent shortages, lack of experience, and lengthy implementation processes. With a global AI talent gap exceeding one million, enterprises need experienced external partners to assist in cost reduction and efficiency improvement, and to minimize pitfalls. Taizhi Cloud has been cultivating the GPU cloud computing power service field for over six years and is one of the few companies in Taiwan capable of providing solutions ranging from AIHPC to generative AI. Furthermore, Taizhi Cloud collaborates with its parent company, ASUS Group, actively developing domestic and international markets, and can bring enterprises the most abundant collaborative resources and solutions.
Taizhi Cloud's AFS POD service integrates one-stop services such as servers, racks, system power cooling, data center network storage, AI computing power resource management, and enterprise application connectivity. It can provide 24/7 enterprise service level (SLA) customer service and technical support, effectively helping enterprises shorten the time and cost of building and maintaining AI data centers.
Many companies collaborate with external partners to develop AI data center projects, which often takes 1-2 years from initial planning to final deployment, resulting in wasted costs due to idle AI equipment. In contrast, Taizhi Cloud possesses extensive experience in system deployment, coupled with ASUS Group's strong support for ASUS server quality, warranty, and customer service. Enterprises can choose NVIDIA, AMD, or other suitable GPUs/CPUs based on their needs, thus reducing AI data center deployment time to as little as 8 weeks, maximizing the effectiveness of their valuable budgets.
While maintaining the stable operation of the AI data center, Taizhi Cloud possesses basic architectural technologies such as liquid cooling and backdoor cooling (RDHE). In response to the large-scale AI and HPC computing needs, it has planned a high-performance storage and IB/ETH high-speed network and platform architecture to meet the needs of large-scale data access while improving power usage efficiency (PUE).
Regarding the cybersecurity protection that enterprises are most concerned about, Taizhi Cloud can also assist in building a security-compliant infrastructure and platform services. In addition to cloud network isolation and virtualization, it can also provide intrusion detection and rapid fault recovery functions. Furthermore, Taizhi Cloud also offers 24/7 customer service and expert technical support services to help users become familiar with how to maintain data center operations.
Customized deployment of large language models accelerates the implementation of AI projects.
In response to the global scramble for AI talent, Taizhi Cloud recognizes that many enterprises lack sufficient AI teams. Therefore, it provides no-code software platforms for both AI 1.0 MLOps development and AI 2.0 generative AI model applications, assisting enterprises in optimizing processes and improving overall operational efficiency. Through Taizhi Cloud's MLOps platform, the difficulty of implementing AI projects can be reduced, allowing teams to quickly bring ideas to fruition, easily manage the AI project lifecycle, and improve team productivity. For cloud service providers' most critical concerns regarding multi-tenancy and resource management, Taizhi Cloud also supports secure multi-node, multi-tenant environments.
"In the area of generative AI services and applications, considering the diverse needs of enterprise applications, Taizhi Cloud can assist users in customizing the deployment of large-scale language models and performing inferences to meet the needs of multiple vertical fields such as smart finance, smart manufacturing, and smart healthcare," explained Li Liguo, Chief Strategy Officer of Taizhi Cloud. "In terms of large language models, we can provide flexible options for different application needs. In addition to the 'Formosa Foundation Model' (FFM) series, which completes the reinforcement training of Traditional Chinese, covering Llama, CodeLlama, and Mistral, there are also mainstream large language models with various parameter levels to meet users' innovative development and accelerate deployment."
AFS POD service's successful integration with the AI ecosystem continues to increase.
Since launching its AFS POD service, Taiwan Intelligent Cloud has helped companies in various industries, including biotechnology, medical care, manufacturing, and finance, build their own big data data centers. For example, in collaboration with ASUS Group, it assisted Chang Gung Memorial Hospital in building an AI inference cloud platform. Currently, it has successfully implemented 12 AI models and LLM generative AI assistants for medical imaging such as X-ray, ultrasound, and EKG electrocardiogram. By using data from the medical image storage system, it has created AI diagnostic assistance, improving the efficiency and accuracy of medical diagnosis and treatment, enhancing the quality of patient care, and improving the doctor-patient relationship.
Meanwhile, Taizhi Cloud is actively cultivating its AI ecosystem, establishing its own AI supercomputing accelerator to provide comprehensive business development resources and technical support for AI startups. In just two years, it has attracted over 60 startups to join, co-creating a wealth of AI application solutions with ecosystem partners. Zhou Binghui, CEO of Taizhi Cloud's AI Supercomputing Accelerator Program, pointed out, "When companies adopt AFS POD services, we can also introduce suitable AI startup teams based on user needs, assisting them in integrating their internal knowledge base using generative AI technology to improve processes and increase operational efficiency. Alternatively, by establishing AI-powered intelligent customer service, we can optimize customer experience and increase sales. Taizhi Cloud will fully act as a bridge for innovative collaboration between companies and startups, accelerating the adoption and transformation of AI to gain a competitive advantage."
Generative AI is rapidly changing the global industrial landscape, and the speed at which enterprises promote AI projects will determine their competitiveness in the market. Taizhi Cloud's AFS POD service not only reduces the total cost of ownership (TCO) for enterprises to build AI data centers, but also improves resource utilization efficiency and reduces maintenance costs. It is also the best solution for empowering employees and customers, optimizing operational processes, and innovating products to achieve enterprise positioning in the market.
[Source:]DIGITIMES】