Hello!
I would like to suggest a voluntary support program for Conan Exiles official servers. This is not a demand, not a complaint, and not a pay-to-win suggestion. The idea is to give players a transparent and optional way to support official server improvements if Funcom considers such a program appropriate.
With Conan Exiles: Enhanced, the game received a major visual and technical upgrade. This update was free for existing players, even though upgrades of this scale are not always free in the game industry. Funcom has already made a major step by releasing Enhanced, moving the game to Unreal Engine 5, improving visuals, and continuing work on fixes, quality-of-life features, and the technical state of the game .
Because of that, I think part of the official server community might be willing to voluntarily support the next practical step: a limited test upgrade of official server hardware. Such a program should be fully voluntary, transparent, cosmetic-only, and officially approved by Funcom and clearly communicated to official-server players before launch.
After Enhanced, many players seem to have returned to official servers. Server load may have increased sharply, especially on older servers with large bases, many building pieces, thralls, pets, placeables, chests, inventories etc. The “Merged Worlds: Exiled Lands & Isle of Siptah” update may also have added pressure, because many Isle of Siptah players moved to merged servers. This may have concentrated more players, bases, inventories, and loading operations on fewer active official worlds, increasing server load.
So some current server problems may be caused by several factors at once: returning players, a major free update, old and heavy worlds, and the Merged Worlds system. Optimization also has limits, especially if parts of the server-side depend heavily on single-thread CPU performance.
Because of this, Funcom could consider something like an Official Server Support Program or Server Upgrade Support Pack for hardware upgrade.
Main idea
Funcom could add a separate Supporter Pack or Supporter tab near Crom Coins purchases. All purchases should be completely voluntary, and all rewards should be cosmetic only, with no gameplay advantage, no priority access, and no VIP status.
Rewards could be similar in quality to popular The Bazaar items: decorative building pieces, weapon and armor skins, placeables, warpaints, or other cosmetics the community actually wants.
Before purchase, there should be a clear confirmation that the player is voluntarily supporting official server improvements. It should also clearly explain that better hardware may improve performance, but cannot guarantee that all lag will disappear, because some problems may be related to game code, network infrastructure, DDoS attacks, world saves, databases, the engine, or other limitations.
Program phases
Phase 1: pilot test on 4 machines
There is no need to raise money for all official servers immediately. The first goal could be funding around 4 high-performance machines (two Intel Core Ultra 9 285K PC and two Ryzen 9 9950X). After that, sales could be paused, and the heaviest PC official servers could be moved there for testing.
Funcom could then compare before/after metrics such as:
- server FPS / tickrate;
- login time;
- base loading time after teleporting;
- chest and inventory opening speed;
- save lag;
- rubberbanding / desync;
- player reports;
- average and peak population;
- RAM usage;
- world save size and total world load.
If the improvement is clear, the next phase could be opened. If the effect is weak, the program could be stopped or adjusted. This would be fairer for both players and Funcom: real data first, scaling later.
Phase 2: Active PC official servers
If Phase 1 works well, Phase 2 could fund around 30 machines and move the most active PC official servers. This might already cover a large part of the real load, because not all official servers are equally active.
Server placement should be dynamic:
- the heaviest PvE / PvE-C / PvP servers — fewer instances per machine;
- medium-load servers — moderate density;
- low-population servers — higher density, if metrics remain good;
- if a server suddenly becomes popular, it can be moved to a less loaded machine during scheduled maintenance.
Low population does not always mean a light server. A PvE server can have only 2–5 players online but still contain huge bases, many chests, placeables, thralls, pets, and a large save file. Placement should consider real world load, not only current online population.
Phase 3: Full PC rollout
If Phase 2 is successful, the program could gradually scale to around 67–70 machines. With a conservative target of about 5 official servers per machine, this would cover roughly 330 servers with spare capacity. If some servers can safely run at 6–10 per machine, the real number of machines may be lower.
Phase 4: console official servers
Console official servers should be considered separately. PC servers are the logical first test platform because Steam/PC statistics and feedback are easier to track. If the PC program works well, the same approach could later be considered for Xbox and PlayStation official servers.
Hardware approach
For dedicated game servers, especially survival and open-world games, total multi-threaded performance is not the only important factor. Strong single-thread performance can be critical. Such servers often have one or several important main threads, so older server CPUs with many weaker cores may perform worse than modern CPUs with stronger single-thread performance. From what I have seen, many server admins for UE4/UE5 or Unity-based games prefer high-end desktop CPUs specifically because they can provide better single-thread performance.
At the same time, the community usually does not know what hardware official servers currently use. If they already run on CPUs similar to Intel Xeon Gold 6349P or other high-frequency server CPUs with good single-thread performance, new hardware may not produce a massive improvement — perhaps only around 8–20% in CPU-limited cases. In that situation, I would not support a full hardware upgrade without a pilot test.
However, even such an improvement may still matter if servers are already near their limit. In game servers, “15% more CPU performance” does not always mean only “15% better gameplay.” If, under heavy load, the server delays or processes some operations less often — for example loading chest contents, inventories, base objects, thralls, pets, or placeables — a small performance reserve may noticeably reduce delays and slow loading.
This is why Phase 1 should be based on real data, not only hardware arguments. If the test shows little improvement, the focus should remain on server-side optimization: saves, networking, inventories, building processing, database operations, and other internal systems. If the improvement is clear, the program can be scaled carefully.
For the heaviest servers, Funcom could test CPUs with very strong single-thread performance, such as Intel Core Ultra 9 285K or similar modern options, or server/platform equivalents with comparable single-thread performance. A conservative test could use around 5 heavy server instances per machine.
For medium-load or low-population servers, a Ryzen 9 9950X or similar 16-core CPU could also be tested at higher density — for example 8–10 servers per machine — if server FPS, save lag, inventory delay, loading time, and RAM usage remain acceptable.
128 GB RAM may be enough for 5 heavy servers, but 8–10 official servers would probably require 192–256 GB RAM, or at least real testing under official-server load.
Rewards
To avoid forcing Funcom to create too much new content, supporter rewards could be smaller than the old Battle Pass. A small thank-you set could include:
- 1–2 base decoration items;
- a unique warpaint;
- a weapon or shield skin;
- an armor cosmetic or appearance item;
- a small supporter badge in the profile, forum, or player list.
The supporter badge should be small and optional. Players should be able to hide their own badge and disable other players’ badges. It should not show the amount contributed and should not create a VIP class. Neutral names like Official Server Supporter or Phase 1 Supporter would be better.
Public metrics
A simple public server status page, or more information in the server list, would also help. Players do not need physical machine CPU/RAM data or sensitive infrastructure details. It would be enough to show:
- online/offline status;
- current population;
- daily peak population;
- status: Good / Moderate / High Load / Critical / Under Review;
- last restart;
- whether the server was moved to upgraded hardware;
- approximate improvement after the upgrade, shown in percentages.
If it is safe and technically possible, Funcom could also show current server FPS or a simple server performance indicator directly in the game, for example in the ESC menu etc. This indicator could be updated in near real time or with a short delay. It would help players understand whether the problem is actually server-side or more likely related to their local client, internet connection, routing, or hardware.
For Phase 1 servers, Funcom could show before/after performance data in the server information panel:
- server FPS: +X%;
- minimum server FPS: +X%;
- inventory delay: −X%;
- login/loading time: −X%;
- restart duration: −X%.
There could also be a Report performance issue button for lag, inventory delay, rubberbanding, login problems, save lag, or server down reports. This would help Funcom identify servers where load suddenly increased.
Why this could help
This approach would give players a voluntary way to support official servers without forcing Funcom to promise a full upgrade immediately. Phase 1 could be paused after the first goal and tested with real data. Rewards would remain cosmetic, and the program would not become pay-to-win.
The main point is not that players must pay for servers, or that normal server maintenance should depend on players. But if part of the community wants to support the next step after the free Conan Exiles: Enhanced upgrade — better official server performance — it would be good to have an official, transparent, and safe way to do it.
