As skyrocketing flash pricing and massive data volumes threaten Artificial Intelligence initiatives, a shift toward autonomous, multi-tier data architectures and DNA-based archiving is emerging to stabilize enterprise infrastructure, as shown at Munich 2026 edition of Technology Live!.
The rapid integration of enterprise Artificial Intelligence (AI) has exposed a fundamental mismatch between modern computational workloads and legacy storage architectures. For decades, corporate data infrastructure accumulated incrementally—built project by project to solve immediate requirements. Today, the concurrent demands of AI performance, cyber resilience, and data sovereignty have overwhelmed these siloed environments. With flash memory prices quadrupling and projected to rise further, the traditional approach of purchasing faster or cheaper point storage products has reached its economic and physical limits, forcing a re-evaluation of data infrastructure operating models.
At the core of the current infrastructure strain is the realization that AI does not represent a single, uniform usecase. Instead, it spans an broad operational spectrum. On one end, running Large Language Models (LLMs) requires extreme GPU utilization and microsecond-latency data delivery; on the other, the long-term retention of massive datasets demands deep, cost-effective archiving. The financial pressure of managing this spectrum has intensified due to severe power limits at data centers and skyrocketing flash memory prices, which have increased fourfold.
To address this convergence of market forces, data infrastructure software provider Scality has introduced its Autonomous Data Infrastructure (ADI) platform. Operating as a software appliance deployed on bare-metal, standard on-premises x86 hardware, Scality ADI establishes a new operating model. Rather than forcing organizations into disruptive hardware refresh cycles, the software-defined stack integrates directly into the Scality operating system to manage data lifecycles autonomously under a single platform and namespace. The architecture manages over 12 exabytes of production data globally, across clients including ten of the top 20 telecommunications companies and seven of the fifteen largest global banks. Individual deployments scale to single-RING failure domains of 250 petabytes of usable capacity, accommodating over six trillion objects across 70 countries.
The technical framework of Scality ADI relies on four primary structural pillars: Multiscale architecture, Cores, AI Connect, and Guardian. The system achieves microsecond latency on object storage, shifting the medium away from its historical classification as purely cold storage. Through proprietary protocols and S3 over Remote Direct Memory Access (RDMA), the platform enables a new “hot tier” that streams data directly from extreme GPU environments to cold storage, maintaining optimal GPU utilization. Conversely, for warm tiers using hard disk drives (HDDs), the system guarantees a throughput of 2 gigabytes per second per petabyte.
A central component of this infrastructure is the Guardian AI co-driver. Operating via a Model Context Protocol (MCP) connection, Guardian utilizes an internal, proprietary LLM to provide agentic operations within the software stack. This framework allows external agents to connect directly to the system, enabling autonomous maintenance, fast recovery, and rigorous control audits. To ensure operational safety, the platform functions with strict structural guardrails rather than acting as an unmonitorable “black box.” Crucially, the system features an “open code” model. While not open-source, this approach allows enterprise clients and service providers to inspect the source code and contribute directly to its development, ensuring a transparent, verifiable, and sovereign trust architecture. The platform’s scaling flexibility extends from “Ateska”—the smallest available storage footprint on the market capable of scaling up to 1 petabyte—to exabyte-scale configurations.
While the immediate demands of AI focus on real-time throughput, the long-term preservation of training data and sovereign workloads presents a separate existential challenge. Traditional media like tape and disk require continuous, energy-intensive refresh cycles. To address the final tier of the data lifecycle, Scality has entered a strategic partnership with Biomemory, an innovator in industrial-grade DNA-based data storage systems. Under the agreement, Scality CEO Jerome Lecat has joined Biomemory’s board of directors to align the technical roadmaps of both entities.
Biomemory, founded in 2021 on research from Sorbonne University and the CNRS French national labs, recently accelerated its commercial readiness by acquiring assets from DNA storage pioneer Catalog Technologies. The partnership aims to integrate DNA storage as a standardized, ultra-secure cold archival tier within modern object storage environments, specifically under the Scality ADI control plane and its CORE5 cyber-resilience framework.
DNA data storage is transitioning from laboratory research into an industrialized IT solution tailored for “keep forever, read rarely” scenarios. The biological medium offers high density, secure immutability, a 150-year retention capability, near-zero media refresh requirements, and zero power consumption when at rest. This infrastructure layer is targeted directly at national archives, genomic repositories, regulated financial or healthcare records, and defense workloads where data loss carries severe legal or societal consequences. Rather than replacing existing flash or disk tiers, DNA storage serves as the logical culmination of a managed multi-tier spectrum, offering an energy-efficient path for mass cold storage that secures data from the API down to the physical architecture.
Dr. Jakob Jung is Editor-in-Chief of Security Storage and Channel Germany. He has been working in IT journalism for more than 20 years. His career includes Computer Reseller News, Heise Resale, Informationweek, Techtarget (storage and data center) and ChannelBiz. He also freelances for numerous IT publications, including Computerwoche, Channelpartner, IT-Business, Storage-Insider and ZDnet. His main topics are channel, storage, security, data center, ERP and CRM.
Contact via Mail: jakob.jung@security-storage-und-channel-germany.de