VK Cloud launches a service with cloud GPU cards for AI and ML scenarios in Kazakhstan - the-tech.kz

VK Cloud announces the availability of graphics accelerators based on the Cloud GPU service in Kazakhstan. The infrastructure is deployed in local data centers, which allows companies to work with personal data in accordance with the Law of the Republic of Kazakhstan No. 94-V “On Personal Data and Their Protection.” VK Cloud provides not just GPU power, but a comprehensive infrastructure for working with data and AI. Within one platform, companies can: - collect and store big data using VK Data Lakehouse - train and run models on the Cloud ML Platform - use GPUs to speed up calculations. This allows you to launch AI projects faster and without the need to assemble infrastructure from different suppliers. Users will be able to analyze big data, develop neural networks and train them in a matter of days instead of weeks. This creates conditions for the emergence of Kazakhstan’s own high-tech products and services based on AI. Scenarios for using GPUs cover a wide range of tasks: - financial organizations can use accelerators to analyze transactions and detect fraud in real time - retail and e-commerce companies will receive tools for personalized recommendations

and Demand Forecasting—In industrial and energy applications, GPUs can help with complex process modeling and maintenance forecasting. The Cloud GPU service will be especially useful for universities, research centers and students working on projects in the field of artificial intelligence. The presence of productive GPUs in local data centers allows you to develop competencies in Kazakhstan while preserving personal data and minimizing infrastructure costs. Startups, in turn, will be able to quickly test hypotheses, train models and enter the market with ready-made solutions, which increases their competitiveness both in Kazakhstan and abroad. Investments in local infrastructure are justified for large enterprises with a constant load, for example, for training neural networks or processing big data. However, in most cases, the most cost-effective way to get GPU servers is to deploy them in the public cloud. The cloud provider provides access to the latest GPUs without the need for recurring capital expenditures. This is especially important given that new processors are released every two years,

and in the near future they will be issued annually. Another benefit of cloud deployment is flexibility. Cloud resources scale almost instantly. It doesn’t matter whether the company uses one server or a multi-node cluster with dozens of GPUs - cloud capacity increases in a matter of minutes. Features of GPU virtualization in the cloud There are several approaches to virtualizing GPUs, each of which has its own characteristics and areas of application. vGPU technology. This approach allows one physical GPU to be shared among multiple virtual machines. Each VM is allocated a certain share of computing power and memory. This technology is used in tasks that do not require the full power of the GPU, or when it is necessary to provide access to graphics accelerators to a large number of users. GPU passthrough technology. This approach provides direct access to the physical GPU directly, without intermediate abstraction layers. This approach provides maximum performance and is designed for tasks that require high graphics performance: from 3D modeling and rendering to cloud computing.

naming. A managed GPU cluster in Kubernetes can operate in several ways: - one pod uses one or more GPUs - resources are distributed across pods via MPS - each process receives allocated resources - resources are shared via MIG - each part runs in isolation. Cloud GPUs provide scalable resources, integration with other services and can become part of a comprehensive infrastructure, rather than simply replacing local hardware. INVITRO is one of the largest networks of medical laboratories in Kazakhstan. Her specialization is laboratory diagnostics and medical services. The company operates in five countries and conducts 260,000 tests daily across 1,930 offices. To speed up service and reduce errors, INVITRO implemented a Face Detection system based on machine learning. The system automatically recognizes faces and objects, comparing them with a database of photographs, eliminating humans from the process and reducing service time. The pilot project was deployed in two medical offices. INVITRO refused local deployment due to the difficulty of operating on a large volume of data. The company chose a cloud deployment integrated with internal corporate services. This

For an hour, cameras in medical offices record and send video to the data center. The algorithm sends the most suitable frames for recognition to the computer vision system, and it recognizes faces and objects based on machine learning from VK Cloud. The Face Detection system has improved the quality of customer service and reduced the time spent searching for a client and registering him. Thanks to this, the registration procedure has been sped up and queues during peak hours have disappeared.