Hi, I'm Glenn Fiedler, I'm a professional game developer working on a 1000 player multiplayer space game.

These days I spend a lot of my time thinking about game server hosting costs, and especially egress bandwidth costs: the cost of bandwidth sent from my server out to players over the public internet.

This is because my space game sends a lot of bandwidth. 10-20 megabits per-second per-client, or 10-20 gigabits per-second for 1000 players.

This is way more than standard low player count games like Apex Legends, Counterstrike, Valorant and Marathon, which usually send something like 256 kbps, 512 kbps or maybe 1mbps per-client.

So when Amazon made this announcement, I was immediately interested:

Starting today, Amazon GameLift Servers provides network bandwidth in and out of AWS at no additional charge for all instance types from generation 6 and later, including On-Demand and Spot, with no commitment required. You now pay only for your Amazon GameLift Servers instance hours; all network bandwidth is free.

Multiplayer game servers generate continuous network traffic to connected players, making bandwidth one of the most unpredictable cost components for game studio customers. With free network bandwidth included, Amazon GameLift Servers eliminates this cost, giving you the simplicity of bare-metal hosting with the global reach of AWS.

Free network bandwidth applies with no enrollment, pricing agreement, or configuration change required. Existing customers on eligible fleets receive the benefit immediately. It is now available in all Amazon GameLift Servers supported regions, except China.

To understand how huge this is, consider a game with an average CCU (concurrent users) of 100,000, with each client sending 1 Mbps. The cost of egress bandwidth at list price from AWS was $1,650,791 per month. Now it's zero.

Of course, nobody would actually spend this much on egress bandwidth. At a scale of 100k average CCU, that's a hugely successful game and they'd have enough clout to be able negotiate a better deal with AWS.

But the reason why this deal is so interesting is that it is democratizing.

Now small indie teams (or even tiny space games) can get access to a deal that makes AWS compute much more attractive for hosting games vs. bare metal, even at list price.

Because of this, I make the following predictions for the next 5 years:

• Without the egress charges keeping small games away from cloud, most new multiplayer games will migrate to AWS GameLift for server hosting from this point forward.
• Bare metal game server hosting companies are in real trouble, being forced now to compete vs. AWS on terms that are much closer to their own costs.
• Google will have to match this deal or give up on the game server hosting vertical entirely.
• More multiplayer games will launch and actually become profitable.
• A totally new class of high player count and high bandwidth multiplayer games are coming as a result of this announcement.

After all, if a 4K video stream costs 25mbps to view and many, or even most of us even can watch these in 2026, why can't games send 25mbps too?

Consider how many more players we can support and how much more detailed our worlds could be if we let go of bandwidth limitations circa late nineties and early 2000s?

Personally, I think it's about time that games start sending more bandwidth, and that's why I think this announcement might just be the best thing that's ever happened for multiplayer games.