Dead Sector

Dead Sector is a single-player data center management game. You start with an empty room and one server you assembled yourself, then build a facility that takes on client workloads. You design the hardware, plan the infrastructure and keep systems running as your operation grows.

Design your servers

Build each server from individual components rather than buying ready-made machines. Processors, memory, storage drives, network cards, GPUs and enterprise AI accelerators each have their own costs, performance and power draw, so you need to match a server to its workload. A machine tuned for log processing, for example, is wasted on an AI training run.

As you expand, you unlock chassis, board form factors and appliances, from compact single-socket nodes to dense clusters combining GPUs, CPUs and storage.

Build the facility

Lay out the floor rack by rack and place your servers. Supply power through UPS units and distribution panels, keeping within circuit limits: an overloaded run can shut down an entire aisle. Connect equipment with network appliances, cables and storage fabrics to create the networks and clusters each workload needs. Facility planning also covers cooling and structured cabling.

Serve clients

Clients bring workloads such as web hosting, video rendering, cold storage, financial trading and cloud gaming. You decide which work your hardware can handle, assign it to suitable servers and earn revenue. Clients can offer quick batch jobs or long-term daily-paying tenancies.

Keeping the facility online means managing failures and costs. Hardware wears out, storage drives fail, and your RAID configuration determines whether a drive failure is minor or causes data loss. Falling behind on the power bill can lead the utility to cut your supply. You can hire technicians to repair and maintain the floor when it becomes too much to manage alone.

Scale up infrastructure

You can connect dozens of racks into a compute cluster for specialized, high-performance workloads that require thousands of CPU cores, or set up GPU clusters for AI training and inference. Pool storage across racks into petabyte-sized systems, configure drive arrays to protect against failures and create replication mirrors for redundancy. Your role is to balance system load, uptime, cost and resilience to keep clients satisfied.

Choose a specialization

A limited research budget and four specialization paths shape how your facility develops: Compute, Storage, AI and Operations. You can focus on raw computing power, become a storage provider, pursue AI workloads or prioritize operational efficiency to make the most of your power and money. Across 100 operator ranks, you unlock new hardware and capabilities.

Manage the operation

You can name servers, query your fleet and use a built-in terminal CLI to issue batch commands. The command-line terminal is optional, and the game can be played at your own pace.