Russia completed deploying a domestically developed data storage system for Moscow State University’s (MSU) supercomputer infrastructure, marking another step in integrating Russian-designed processors, server hardware, and software into high-performance computing (HPC). The SDS TROK software-defined storage platform, developed by the Astra Group, integrates with servers manufactured by Spintronika and powered by 48-core Baikal-S processors.
The installation forms part of MSU’s Istok hardware-software complex, which has been operating since 2026. The project was officially announced on November 9, 2026, following the completion of development and integration work by Astra Group and Spintronika, with financial support from Russia’s Ministry of Science and Higher Education. The system offers storage infrastructure for scientific computing workloads and enables researchers to access stored data via the Network File System (NFS) protocol.
The project is significant because the university specifically required compatibility with Russian processors and the use of domestically developed components. The participating companies prioritized establishing compatibility between the software and Baikal-S-based server hardware, rather than adapting an existing foreign storage platform.
SDS TROK: A Storage Platform Built Around Russian Hardware
Astra Group has developed a software-defined storage platform known as SDS TROK. Unlike traditional storage systems, which rely on dedicated proprietary hardware, software-defined storage manages storage resources distributed across multiple servers using software. This method enables administrators to consolidate hardware resources into a centrally managed storage environment and increase capacity by incorporating additional nodes.
SDS TROK was implemented on servers developed and manufactured by Spintronika using Baikal-S processors for the MSU project. Users access shared data through NFS, a standard network file-sharing protocol extensively used in Linux environments, as the system provides storage services to the Istok computing complex. Researchers can use the storage infrastructure without the need to implement an alternative data-access procedure as a result of this arrangement.
The Ministry of Science and Higher Education provided support for the initiative. MSU implemented a competitive program to secure funding for modernizing the material and technical infrastructure of organizations that conduct scientific research and development.
The integration process was a critical component of the project. A dedicated software build and assembly pipeline was established, followed by two phases of testing. The first stage involved validating software compatibility using virtual machines, while the subsequent stage involved testing physical servers equipped with Baikal-S processors. The developers reported that both stages exhibited stable operation and adherence to the necessary specifications.
The main SDS TROK codebase has also been updated to incorporate the modifications developed for Baikal-S compatibility. This implies that the project is not limited to the MSU installation. Compatibility enhancements are also available to customers deploying the storage platform on Russian processor architectures.
Baikal-S: The 48-Core Processor Behind the System
The basic hardware component of the storage servers is the Baikal-S, also known as the BE-S1000. Baikal Electronics created it, and it is intended for use in storage infrastructure, high-performance computation systems, and server platforms.
The processor implements the Armv8-A architecture, which includes 48 ARM Cortex-A75 cores and operates at a frequency of 2.5 GHz. It has a maximum capacity of 768 GB per processor and supports six DDR4-3200 channels with ECC functionality in its memory subsystem. The processor can also connect high-speed storage controllers and other peripheral equipment, as it offers up to 80 PCIe 4.0 lanes and 32 MB of L4 cache.
The specified thermal design power is approximately 120 watts. These attributes render Baikal-S especially pertinent to infrastructure projects that necessitate significant memory bandwidth, connectivity, and parallel processing capabilities.
Processor compatibility is just one aspect of the engineering challenge for a storage platform. The hardware must also support the operating system, storage-management software, network interfaces, and data replication mechanisms. Consequently, the MSU deployment illustrates the processor’s ability to function as a component of a comprehensive infrastructure, rather than a standalone server.
DRBD Replication Protects Scientific Data
The SDS TROK installation’s data protection strategy is one of its most critical attributes. Frequently, scientific computing systems execute calculations for extended periods. The loss of stored results due to a server failure can waste significant computing time and necessitate that researchers perform additional calculations.
SDS TROK resolves this issue by using Distributed Replicated Block Device (DRBD) technology to implement synchronous replication. A write operation is only acknowledged under this arrangement after the data has been stored on the necessary number of nodes. In the event of a server failure, the replicated information is still accessible on the other participating nodes.
The system also includes automatic recovery procedures. Delta synchronization is initiated by the storage cluster upon the return of an unsuccessful server to operation. Instead of retransmitting the entire dataset, it copies only the blocks that changed during the interruption. This reduces network traffic and enables the affected node to revert to the cluster’s fault-tolerant configuration more efficiently.
Another component is the division of responsibilities among various node categories. The configuration and overall coordination of the storage environment are managed by controllers, while DRBD resources are managed by worker nodes. This architecture enables the system to expand while simultaneously ensuring centralized administration.
According to the developers, administrators are not required to manually manage replication or initiate recovery procedures after each failure. Interruptions should be minimized from the perspective of researchers accessing the NFS volumes, while previously stored calculation results should be safeguarded.
Baikal-S Expands Its Role in Russian Infrastructure
The MSU installation is a component of a more comprehensive initiative to establish Baikal-S as a platform for Russian server and infrastructure equipment. Several domestic hardware and software developers have already expressed interest in its 48-core architecture.
For instance, Aquarius has provided the S2-1 storage infrastructure, built on its AQ2BS1000 server and equipped with Baikal-S. The system is designed for protected IT infrastructure, which includes government organizations and critical information infrastructure.
Another example of the platform’s application in domestic computing infrastructure is Graviton’s S2124B server, which is outfitted with two 48-core Baikal-S processors. The system is designed for high-performance computing and artificial intelligence applications.
Astra Group has also collaborated with Baikal-S in other sectors of its product line. Astra Cloud was commercially launched on the same processor architecture, while the Tantor XData 2B database machine was adapted for the company’s internal 1C environment.
Istok Moves Towards Full-Scale Scientific Computing
Since 2026, the Istok complex has been operational at Moscow State University. It is utilized by research teams from MSU and over ten organizations affiliated with institutes of the Russian Academy of Sciences.
The system is designed for computationally intensive research, such as environmental studies, biomedical research, cognitive science, and climate modeling. Throughout the computing process, these applications generate substantial amounts of data that must be stored, accessed, and safeguarded.
The university’s preliminary performance evaluations suggest that climate-modelling tasks are being completed at a significantly faster pace than they were on computing resources of comparable size that were previously employed. Beginning in 2027, the university intends to transition to full-scale scientific calculations and research projects.
Therefore, the storage system will assume a critical function as the workload increases. Large-scale scientific simulations require more than just processor performance. Additionally, they require the capacity to preserve intermediate and final results, as well as reliable storage and adequate network bandwidth. The efficiency of the complete computing operation can be impacted by a failure in any of these areas.
Russia’s HPC Independence: A Test Case
The SDS TROK deployment’s final outcome is a major step in Russia’s quest to establish a comprehensive domestic computing ecosystem. In an operational scientific environment, the project integrates Astra Group’s storage software, Spintronika server hardware, and Baikal-S processors. Its value is not limited to the replacement of imported storage equipment; rather, it is rooted in the demonstration that Russian-developed technologies can be integrated, tested, and deployed in conjunction for rigorous research applications. The platform’s reliability and scalability will be further tested by the anticipated increase in Istok’s scientific output beginning in 2027.