Listen to this article:
From storage to intelligent platform: What's changing in DSM?
For decades, Synology DiskStation Manager (DSM) has been the standard for data management, backup, and file sharing. With the new announcement, however, the company is taking a path that puts it in direct competition with cloud giants. Instead of users having to "pour" their data into external services to gain the benefits of artificial intelligence, Synology wants to bring intelligence to the data.
According to information from the official announcement for the enterprise segment, the new generation of DSM focuses on creating private knowledge bases. This means that your business documents, system logs, and metrics will become the foundation for AI agents working within the closed environment of your local server.
What does "Agentic AI" mean in practice?
The term agentic AI has been very popular in the tech world in recent months. Unlike ordinary chatbots, such as standard ChatGPT, which only answer queries, AI agents can perform actions. They can plan steps, use tools, and complete complex tasks.
In the context of Synology, this means that the DSM Agent won't just be a "smarter search engine." For example, it can:
- Automatically analyze system logs and suggest configuration fixes.
- Organize and sort thousands of documents based on their content.
- Orchestrate workflows between different applications within the Synology Office Suite.
Privacy vs. Cloud: Why is local inference key?
Today's businesses face a huge dilemma. Using models like GPT-4 or Gemini increases productivity, but at the same time poses a massive risk to the protection of trade secrets and regulatory compliance. Once data leaves your infrastructure, you lose full control over it.
Synology solves this problem using on-premises inference. This means the model's "thinking" itself takes place on your hardware. To achieve sufficient performance, Synology plans to use its GPU rack servers and specialized AI appliances.
Comparison of approaches:
| Feature | Cloud AI (e.g. ChatGPT) | Synology Private AI |
|---|---|---|
| Data privacy | Data leaves to the provider | Data stays with you (on-prem) |
| Costs | Subscription per user/tokens | Investment in own HW |
| Regulations (GDPR/EU AI Act) | Complex compliance necessary | Easier compliance thanks to data control |
Impact on the Czech and European market
For Czech companies, especially those working with sensitive information (lawyers, doctors, public administration, or technology startups), this announcement is significant. The European Union, through the EU AI Act and strict GDPR rules, places great emphasis on how data is processed during AI training and use.
The ability to run AI locally, directly in the office or in a company data center, eliminates the need for complex legal agreements on personal data processing with American cloud providers. For the Czech market, this means that even smaller businesses can adopt advanced AI technologies without fear of violating legislation.
Management at scale: Cluster Manager and Active Insight
Synology isn't just addressing individual units, but also large corporate networks. The new Cluster Manager will allow administrators to manage an entire fleet of NAS devices from a single interface. Thanks to application and service containerization, workloads can now be flexibly moved between individual servers.
Furthermore, the Mass Deployment feature within the Active Insight tool will enable rapid configuration of new devices, which is critical for IT departments that need to scale their infrastructure without manually setting up each individual server.
Price and availability
Synology has not yet announced an exact price for the software updates themselves (DSM is usually free for existing users), but emphasis is placed on the hardware. A cheap home NAS won't suffice for effective running of agentic AI workflows. Companies will likely need models with higher processor performance or dedicated units with GPU support.
We expect that specialized AI servers from Synology will have a price level in the range of tens to hundreds of thousands of Czech crowns, which corresponds to their enterprise focus. For regular users, a more affordable version with limited performance will be available, using the CPU for simpler tasks.
Will I need to pay a monthly subscription for running AI in Synology, like with ChatGPT?
The main advantage of Synology is precisely that it focuses on on-premises solutions. This means you don't pay for tokens or a monthly cloud service subscription, but invest in your own hardware that runs AI locally.
Can Synology DSM run on my existing NAS device?
That depends on the hardware specifications. Agentic AI and local inference require significant computing power (especially RAM and ideally a GPU). Older models with low performance likely won't support this advanced feature or will have very limited support for it.
How secure is this solution from a cybersecurity perspective?
It is significantly more secure than cloud alternatives because the data never leaves your corporate network. However, if your NAS is not properly secured (firewalls, updates, access rights), it could be exploited as an entry point into the network.
