The short version
Choose US East (Ohio) for audiences on the East Coast, in the Midwest, in eastern Canada, or when you also serve Europe from the same server. Choose US West (California) for the Pacific coast, the Mountain states, western Canada, or users in Asia-Pacific.
Side by side
| US East | US West | |
|---|---|---|
| Location | Ohio | California |
| Processor | Intel Xeon Gold 6248 | Intel Xeon Gold 6248 |
| Network | 1 Gbps unmetered | 1 Gbps unmetered |
| Price | From $5/mo | From $5/mo |
| Closest to | East Coast, Midwest, eastern Canada, Europe | West Coast, western Canada, Asia-Pacific |
Hardware and pricing are identical, so this is purely a question of geography.
Thinking about latency
Ohio sits within a short hop of the large East Coast population centres and the Chicago network hub, which makes it a balanced choice for the eastern half of North America. It is also the better US region if a meaningful share of traffic comes from Europe, since transatlantic routes land on the East Coast.
California is the natural home for users west of the Rockies and the shortest path from our US regions to Japan, Korea, Australia and Southeast Asia, because transpacific cables land on the West Coast.
Latency-sensitive workloads
For game servers, voice chat, trading tools and interactive apps, the difference between 20 ms and 70 ms is noticeable. Put the server on the side of the country where most players or users are, and if they are evenly split, consider one server per coast. For APIs and websites behind a CDN, the origin location matters less, and US East is often the default because of its central position.
Running both coasts
A dual-region setup is straightforward: deploy the same image in Ohio and California, keep data in sync with database replication or scheduled backups, and route users with DNS-based geographic steering or a load balancer. Each server has its own dedicated IPv4, unmetered transfer and the same control panel, so the second region is just another server in the same account.
How to decide with data
Check where your current visitors connect from in your analytics, then measure. Deploy a Micro server in each region for a day and test from the networks your users are on with mtr, ping or a browser-based latency checker. Keep the winner and delete the other; there are no setup fees to recover.