Improve a Valheim server by identifying the actual bottleneck before changing hardware. Separate client-frame-rate problems from server simulation, network latency, save-time hitching, and mod incompatibilities; then change one variable at a time. A dedicated server can move persistent-world hosting away from a player’s gaming PC, but it cannot repair a weak client GPU or a poor internet connection.
Playing with friends? Godlike.Host runs hosted Valheim servers from 10 data-center locations, with free DDoS protection.
What “server performance” means in Valheim
Players use “lag” to describe several different failures. The remedy depends on which one you are seeing. Record the symptom, who experiences it, and when it occurs before making changes.
As an Amazon Associate I earn from qualifying purchases.
Quick wins for a faster PC:
Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Clear out junk files and repair common Windows errorsFree Scan →| Symptom | What players notice | First area to inspect |
|---|---|---|
| Client frame-rate drop | The game renders slowly on one computer while other players remain responsive. | That player’s graphics settings, GPU, CPU load, and background applications. Moving the world to a dedicated server will not raise the client’s rendering performance. |
| Rubber-banding or desynchronisation | A character snaps back, creatures teleport, doors react late, or movement pauses and catches up. | Latency, packet loss, Wi-Fi, tethering, competing uploads, and the server’s CPU or memory pressure. |
| Delayed interactions | Mining, building, combat, or inventory actions take several seconds to register for several players. | Server simulation load, network upload, and process logs while the delay is reproduced. |
| Save hitch | The whole group pauses at regular intervals, often during an automatic save. | Disk activity, world complexity, save interval, and backup work. |
| Whole-party slowdown | Everyone sees the same stalls at once, often after a world has grown or a mod was added. | Server resource usage, world size and complexity, mod versions, and the network path shared by all players. |
A useful test is to ask every player to describe the same event using one of these categories. If only one person has low frame rates, do not treat it as a server fault. If everyone sees delayed actions at the same moment, collect server and network measurements during that event.
What a dedicated server changes—and what it cannot fix
Iron Gate describes a dedicated server as a way to provide a persistent world that friends can join at any time. The world process runs separately from a player’s game client, so the person who normally hosts no longer has to render the game and host the simulation on the same computer. That separation can make a busy home-hosting arrangement more predictable and allows the world to remain available when the usual host is offline.
It does not make every multiplayer problem disappear. Players still need a reliable route to the server, and each client still needs enough CPU and graphics capacity to render the scene. A dedicated process can also be under-provisioned, misconfigured, or affected by an incompatible mod. Treat it as a change in architecture, not as an automatic performance upgrade.
Step 1: Establish the symptom before changing anything
Create a short incident record
For each episode, note the clock time, the players affected, the activity in progress, and whether the problem clears by itself. Record whether the world was saving, whether a new region was being explored, and whether anyone had just joined. This gives you a repeatable event against which to compare a configuration change.
The Tool Desk
Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →- Mark the issue as client FPS, rubber-banding, delayed interaction, save hitch, or whole-party slowdown.
- Write down whether the affected player is on wired Ethernet, Wi-Fi, or a mobile tether.
- Note any simultaneous downloads, cloud synchronisation, streaming, or backups on the host network.
- Record the server’s CPU, memory, disk activity, and network upload at the time of the event.
- Keep the Valheim version, server backend, BepInEx version, and enabled mod list with the incident.
Use a controlled reproduction
Ask the same group to repeat a troublesome activity after each change. Do not change the save interval, hardware allocation, network route, and mods simultaneously; otherwise you will not know which action mattered. If the symptom disappears, repeat the test once more to distinguish a real improvement from a temporary quiet period.
Step 2: Verify the server’s network path
Steam backend
Steam-backend servers use UDP ports 2456 and 2457 by default. Forward those ports from the router to the server host and allow them through the host firewall. Verify that the forwarding rule points to the server’s current local address; a changed DHCP address can silently invalidate a rule. Test from outside the local network, because a local connection does not prove that the public path works.
Crossplay backend
The Crossplay backend sends server traffic through a relay. External players therefore do not need router port forwarding to reach the server from outside the local network. This removes one common administration failure, but it does not guarantee low latency: the relay and each player’s internet route still matter.
Running more than one instance
Do not casually reuse the same public ports when several servers share one public IP address. Give each instance a distinct port range and a distinct instance identity wherever the server configuration requires one. Keep a written map of instance, internal host, public ports, save location, and firewall rule so that troubleshooting one world does not disrupt another.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Rank #2
- Viking Game laminated laptop stickers All approx 3"-4" in length
- Laminated, Not UV Proof. Will fade in sun. Best used indoors.
- Great for Laptops, Phones, Computers, Tablets and more!
- Handmade by a Stay at Home Small Business Mama!
| Backend | Default network detail | Administrative implication |
|---|---|---|
| Steam | UDP 2456–2457 by default | Forward the ports to the server and permit them in the firewall. |
| Crossplay | Uses a relay for external access | No router port forwarding is required for players outside the LAN, but route quality still affects responsiveness. |
Step 3: Measure the bottleneck while the problem is happening
Upgrading a component that is sitting idle will not cure rubber-banding. Watch the server process and the host system during a reproduction rather than relying on an average from a quiet session.
CPU
Check total CPU usage and, where your operating system provides it, per-core utilisation for the server process. A consistently busy core while interactions stall points toward simulation capacity. A brief spike during a save is different from a sustained ceiling during normal play; record both patterns.
Memory
Track used memory, available memory, and swap or paging activity. The community guidance for a small Linux VPS uses 2 vCPU and 4 GB of RAM as a starting consensus, not an official minimum. It also warns that memory use can rise as the save becomes more complex. If the process is paging or the operating system is reclaiming memory during the incident, increase the available memory only after confirming that this is the observed constraint.
Disk activity
Watch write latency and queue depth while a save or backup runs. A fast-looking storage device does not help if another service is saturating it at the same time. Keep the world and its backups on storage with enough free space, and avoid running large downloads or scans during your busiest play period.
Free tools Windows power users keep installed
One-click scans. No signup required.
Network upload and packet loss
Measure the server’s outbound traffic and look for a saturated uplink. A connection can show a high headline speed yet suffer latency spikes when uploads fill the queue. If server CPU and memory remain comfortable while players rubber-band, test packet loss and route quality before buying RAM or changing the processor.
Process logs
Capture the server log around the timestamp of the incident. Look for repeated connection errors, save warnings, mod exceptions, or version messages. A single warning is not proof of the cause; correlate it with the player-visible symptom and the resource graphs.
Rank #3
- Gaming Inspired Design
- Geek Culture
- Lightweight, Classic fit, Double-needle sleeve and bottom hem
Step 4: Remove common network interference
Temporarily eliminate competing traffic so that you can tell whether the game is fighting for the connection. Pause torrents, cloud-sync uploads, operating-system downloads, game updates, and video streams on the host network. If the issue vanishes, re-enable services one at a time to find the source and schedule or rate-limit it outside play sessions.
Use wired Ethernet for the server whenever practical. Players who experience repeated rubber-banding should test a wired connection as well, especially when their Wi-Fi signal crosses several rooms or shares a crowded channel. Mobile tethering can introduce variable latency and carrier-side congestion; compare it with a fixed connection before changing server hardware.
PC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchTraffic-shaping, packet-filtering, or other utilities that modify network traffic can interfere with games. Disable such software briefly for a controlled test, or place the game and server traffic outside its management rules. Restore normal security controls after testing and use the result only as evidence about the network path.
Step 5: Choose hardware from measurements, not a generic specification
What is officially known
The reviewed official material does not publish a minimum dedicated-server specification. That means a number copied from a hosting advertisement should not be treated as a universal requirement. World complexity, concurrent players, mods, operating-system overhead, and other services all change the load.
A cautious starting point for a small VPS
A community guide gives 2 vCPU and 4 GB of RAM as a consensus starting estimate for a small Linux VPS. It is explicitly a starting point rather than an Iron Gate requirement. Begin there only for a modest world, monitor it during real play, and leave headroom for the operating system and any additional services. Expect memory demand to increase as the save becomes more complex.
When more resources are justified
- Increase CPU capacity when a core remains saturated during simulation and players report delayed actions.
- Increase RAM when available memory collapses, paging begins, or the server is killed under load.
- Improve storage or isolate the world when save-time disk latency coincides with the hitch.
- Improve the uplink or remove competing traffic when network queues fill while CPU and memory remain normal.
After a change, repeat the same reproduction and keep the old measurements. If the graphs do not show a previously constrained resource, roll the change back and investigate the network, save process, or software versions instead.
Rank #4
- Valheim Boar Loves You quote. A perfect design for fans of the best selling Viking game.
- An amazing and unique design which is also perfect as a gift.
- Lightweight, Classic fit, Double-needle sleeve and bottom hem
Step 6: Tune saves and backups carefully
Understand the documented defaults
Iron Gate’s guide documents a default save interval of 1,800 seconds, or 30 minutes. It also documents four automatic backups, a 7,200-second short-backup interval, and 43,200-second intervals for subsequent long backups. These values describe the server’s documented behavior; your launch configuration should be checked after every update.
| Setting | Documented value | Operational meaning |
|---|---|---|
-saveinterval |
1800 seconds |
Automatic save every 30 minutes by default. |
-backups |
4 |
Keep four automatic backups. |
-backupshort |
7200 seconds |
Short-backup interval of two hours. |
-backuplong |
43200 seconds |
Later backup interval of twelve hours. |
Changing the save interval
A shorter interval limits the amount of progress that could be lost after a crash, but it can increase save activity. A longer interval reduces save frequency but leaves more recent progress at risk. Change only the save-related argument, reproduce the workload, and watch disk activity and hitch duration. Keep the setting that matches your world’s recovery tolerance rather than choosing the shortest possible interval automatically.
Protect and test restoration
Retain multiple backups and copy them to storage that is not affected by the same disk failure as the live world. Test restoration with the server stopped: identify a backup, restore it to a separate location, start a test instance, and confirm that the expected world and player data load. A backup that has never been restored is an assumption, not a recovery plan.
Step 7: Audit mods and versions
Mods can add useful controls, but they also introduce another compatibility boundary. Keep the server and every client on compatible Valheim and BepInEx versions. When a problem starts immediately after an update, compare the version and mod list with the last known-good profile before changing hardware.
Dedicated Simulation
The Dedicated Simulation listing documents queue and target-FPS controls. It requires BepInExPack_Valheim 5.4.2350 or newer according to that listing. Treat those requirements as a compatibility statement, not as proof that every world will run faster. Back up the world, change one setting, and observe the same reproduction.
SmoothServer
SmoothServer’s repository describes adaptive budgets, asynchronous saving, and telemetry, and states a target of Valheim 1.0.14 with network version 40 and BepInEx 5.4.2350. Those declarations identify the versions the project targets; they do not guarantee an improvement on a different Valheim build, world, or mod combination.
Best Value
- Gaming Inspired Design
- Gaming and Geek Inspired
- Lightweight, Classic fit, Double-needle sleeve and bottom hem
A safe mod workflow
- Stop the server and make a verified world backup.
- Record the current Valheim, BepInEx, server, and client versions.
- Install or update one mod only, using the version it documents.
- Start a test instance or a copied world before touching the live save.
- Reproduce the original symptom and inspect logs, CPU, memory, disk, and network data.
- Keep a rollback profile with the previous files and launch settings.
If removing a mod resolves the issue, report the exact versions and reproduction to its maintainer instead of assuming that a different tuning value will work universally.
Home hosting, a VPS, and managed hosting
Choose a hosting model by the problem you are trying to solve, not by the label “dedicated.” Compare who supplies compute and uptime, how players reach the server, how much control you have over saves and mods, and how much troubleshooting you must perform yourself.
| Option | Compute and uptime | Network administration | Control and troubleshooting |
|---|---|---|---|
| Home server | You supply the computer, power, cooling, and availability. | Steam backend normally requires forwarding UDP 2456–2457; Crossplay uses a relay for outside-LAN access. | Maximum access to logs, files, backups, and mods, but you own every diagnosis. |
| Linux VPS | A provider supplies virtualised compute; you still choose capacity and monitor contention. | You configure the firewall and server path, then verify latency from players’ locations. | Good shell and file access, with responsibility for updates, backups, and recovery. |
| Managed Valheim server hosting | The provider handles the persistent service and usually its underlying availability. | The service may reduce router work, but you should still check its routing and player geography. | Usually simpler controls and support, with less control over the host operating system and underlying diagnostics. |
Move from home hosting to a managed service when administration time, uptime, or network reachability is the limiting factor. Move to a larger VPS or home machine only when measurements show CPU, memory, disk, or uplink pressure. Neither move is justified solely by a client-side FPS problem.
Troubleshooting branches
Everyone rubber-bands, but the server is not busy
Test packet loss and latency from several players, pause all competing uploads, and compare wired Ethernet with Wi-Fi or tethering. Check the selected backend and the public route. If the problem follows one player rather than the group, focus on that player’s connection instead of the server.
The pause coincides with a save
Confirm disk activity and the save timestamp in the log. Keep backups, then test a different save interval one change at a time. If the pause remains despite low disk utilisation, inspect mods and process errors rather than repeatedly shortening the interval.
Recommended Free Tools
Performance degraded as the world grew
Compare current memory and CPU traces with earlier incidents. The community guidance specifically warns that memory use can rise with save complexity. Add resources only when the graphs show pressure, and check that no background service is consuming the remaining headroom.
The issue began after a mod or game update
Return to the last known-good mod and version profile on a copied world. Verify the compatibility targets, including the BepInEx requirement stated by the mod, then reintroduce changes individually. Keep the live world backed up and do not test an unverified combination directly on it.
A practical maintenance checklist
- Keep a written record of backend, public ports, firewall rules, instance identity, and save location.
- Check CPU, memory, disk, upload, and logs during a normal busy session, not only when the server is idle.
- Schedule downloads, cloud synchronisation, and backups outside the busiest play periods.
- Verify that automatic backups exist and periodically perform a stopped-server restoration test.
- Document Valheim, BepInEx, and mod versions before every update.
- Change one variable at a time and repeat the original reproduction.
- Keep a rollback profile for launch arguments, mods, and the previous world backup.
The most reliable performance improvement is a shorter feedback loop: identify the symptom, measure the relevant resource or route, remove interference, make one controlled change, and test again. That process prevents an expensive hardware purchase from masking a port, Wi-Fi, save, or compatibility problem.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
