Can not recommend this game. This game came off as a great game. But upon playing it you start to notice things that seem out of place. The game was released into 1.0 and not early access. Sure the game is 10$ but this does not feel like 10 dollars.

---Pros---
1. Cool cable routing system.
2. Cool packet simulation system.
3. Cool and functional rack mounting system.
4. LAG (technology) is built into the switches.
---Cons---
1. No VLANs. Everything just doesnt talk to anything else magically.
2. Money income is slow and honestly feels like an idle game.
3. No routing or any other technologies. Only the arbitrary IOPS that are usually used to calculate speeds of storage.
4. The in game tutorial...... Are these written by AI? Were these not checked for what the game uses? We are talking about IPv6. Front-Back cooling as if we have that, Server weights as if thats in the game, among many other things.
5. The servers are all the same except slightly retextured.
6. The switches are made for a spline setup but again no vlans so it just magically works aggregating everything.

This just seems like an idle game. Again Packet simulation super cool. Cable management super cool. The content of the game is severely lacking for a 1.0 release.

----Edits for clarity below----
The game really does not give us any concepts that are worth while. It talks about Spline-Leaf Architecture but as many have pointed out does not balance or do anything with the Spline-Leaf that is expected. VLANs are not a thing so keeping everyone seperate into only their own racks makes the most sense. Its not like we are using 2 WAN links from our carrier no we are only managing their internal ports. Their internal ports are using external IP addresses. We do not assign these addresses to servers except in extremely rare cases. The main thing this game is good at is the cable management. For education I would not recommend this.

A great idea buried under an unfinished game

The concept here is genuinely exciting: build a datacenter from scratch, physically rack your servers, run cables by hand, and watch network traffic visualize in real time. If that pitch speaks to you, you are not alone. It spoke to me too. Unfortunately, the execution falls well short of what the idea deserves, and what is here feels closer to an early alpha than a finished product.

From a networking standpoint, the simulation is extremely thin. There are no VLANs, no real switching logic, no routing, no DHCP, no per-interface configuration, and no separation between networks. The highest port-count switch available tops out at 32 ports — for context, standard datacenter switches begin at 288 and can exceed 1,000. Bandwidth is technically presented as a constraint, but in practice a single rack will never come close to saturating it, making the whole concern moot. Without meaningful bandwidth limits, higher port-count options, or actual networking concepts to engage with, the design space collapses quickly.

The hardware layer suffers from the same shallowness. Servers have no configurable interfaces, no SFP slots, and no architectural trade-offs — they are essentially IOPS generators you place and forget. There is no virtualization layer, no hypervisor model, no overcommit or resource pooling of any kind. Each client gets dedicated hardware, which means "scaling" is just adding more of the same rack over and over. That is not datacenter design; it is copy-paste. The fact that the game also lacks any templating system makes this repetition genuinely tedious.

Client management barely exists. Once a client is acquired, they become a static fixture. You cannot reallocate resources, adjust services, manage SLAs, or evolve their setup in any meaningful way. This strips out one of the most interesting parts of running infrastructure.

The cabling system is the game's strongest idea, and it still feels unfinished. Cable supports are misaligned and visually inconsistent, and small interactions — like adding SFPs one at a time to a 32-port switch, 64 times over — become a chore rather than satisfying fiddly work. The UI is dated, the office environment feels largely empty and autogenerated, and the tutorial fails to explain even basic concepts clearly.

The result is a game that does not know its audience. It is too shallow for anyone with real infrastructure knowledge, and too poorly explained for someone trying to learn. The idea is excellent. The game, in its current state, is not. Come back in a year or two — but do not buy it today.

Score - Refunded 1.5/10

I struggle to recommend this for a variety of reasons.

For a "sim" it is simplistic. Switches are low port counts for what you'd expect in a data center, you're locked into top of rack design because of it, which would be fine, if each customer wasn't its own WAN as well. Servers are a fixed bandwidth, I was disappointed to learn that not a single server accepted a sfp. The tutorial tries to indicate to you that bandwidth is a concern but you literally never have to worry about it, a single rack will never exceed 3.75 gib/s. Essentially you template 1 rack with 4 switches, and N racks with two switches, but then the game doesn't let you template things so it becomes monotonous to buy and just repeat the same design for each client.

Without deeper networking concepts, reducing the number of WAN connections, higher port count switches, and genuine bandwidth constraints, I can't recommend this to anyone. It's insane that the highest port count switch you have available to you is 32 ports, your standard switch in the real world is 24 or 48 ports and data centers regularly make use of large rack sized switches that *start* at 288 ports (fun fact, they can go up to 1k+ ports).

As it is today, its not a data center sim, its a cabling sim that appeases your OCD. Having to grab sfps one at a time is silly, I should just be able to hold the box in my hand and add/remove sfps. Have fun repeating that 64 times for a 32 port switch... I expect a degree of monotony from any sim, this game goes above and beyond throwing monotony at you and at the same time strips everything that makes data centers interesting from it.

Mandatory edit: developer quickly made it so that you can insert SFPs from the box. Props for that change.