Summary
- Scaleway plans three new Stockholm zones and three new Frankfurt zones in 2027.
- The six proposed availability zones are separate from its operating three-zone Milan cloud region.
- The Swedish AMD and Danish NVIDIA GPU proposals are different facilities, not additional operating availability zones.
Scaleway intends to open six additional cloud availability zones in Europe during 2027, with three planned in Stockholm and three in Frankfurt. The company announced the roadmap at its ai-PULSE event on 13 October alongside its opening of two more Milan zones and separate plans for dedicated Nordic GPU systems.
All six Swedish and German zones are planned for 2027 and are not presented as already operating. Scaleway has not identified completed buildings, confirmed construction status or electricity capacity for them in the announcement. Their intended opening year should therefore not be treated as an assured date for commercial availability.
A cloud region consisting of multiple independent zones is designed to let customers place workloads across different physical sites. The architecture can reduce exposure to faults affecting one building or one local infrastructure group. The result depends on application design, with data replication, load balancing and failure recovery necessary before a multi-zone region can be considered resilient in practice.
Separating the roadmap from available services
Scaleway says its Milan region is now operating across three zones. The Stockholm and Frankfurt plans are different: they have not yet reached the same operating milestone, and service availability details may change as projects proceed.
Planning three zones in a metropolitan area imposes commercial and engineering requirements beyond leasing a single server hall. Sites need suitable power supplies, network paths and inter-site communications. At the software layer, region-wide services require supporting control systems and consistent customer-facing interfaces. None of those arrangements has been quantified for the two planned regions.
The company already operates European cloud infrastructure and positions the expansion as a way to give customers more geographic choice. Such choice may support data-residency preferences, but regulatory compliance cannot be inferred solely from the location of a data hall. Customers also need to understand where backups, telemetry, administrative access and support processes operate.
Stockholm and Frankfurt are established connectivity and colocation markets with competitive demand for power and land. The physical feasibility of a new availability zone depends on the specific facility, contracted electricity supply and available network connections. Scaleway has not published the dates at which the planned sites will receive power or begin accepting production applications.
The group also plans Swedish AMD and Danish NVIDIA compute systems, described as dedicated GPU Zones. Those services address accelerator-intensive workloads and should not be counted as additional conventional availability zones unless the provider explicitly says they deliver that architecture.
Developers and cloud buyers will need to separate regional branding from the physical location of replicas and backups. Two sites within the same metropolitan area may share transmission corridors, utility assets or upstream dependencies even if they occupy different buildings. Conversely, distributing work across countries may introduce longer data paths or regulatory questions. Scaleway has not released a common-failure assessment for the planned six zones. The announced number of locations offers a roadmap for procurement, rather than a completed resilience certification.
The six zones currently represent a 2027 deployment roadmap rather than completed capacity. Scaleway has not published their individual facility locations, electricity contracts, commissioning milestones or full service inventories.
