Full Turbo Dedicated Bare Metal Server

To get started with ERPC Bare Metal, create an account in the ERPC Web Dashboard at dashboard.erpc.global, then choose a Bare Metal plan and region. This reserves a physically dedicated server for your use alone, ready for your Solana RPC workload. Pay by card, or use SOL, USDC, or EURC to buy ERPC Credits and activate your plan.

If you are competing to land first on Solana slots, choose Bare Metal. It is a physically dedicated machine reserved for you alone, so no other user can take your resources. A VPS is a virtual environment that divides physical CPU cores into multiple vCPUs, which is great for cost-performance but introduces the contention a slot race cannot afford. Pick Bare Metal for maximum performance, and a VPS when cost-performance matters more than the last millisecond.

A dedicated endpoint makes you the sole user of the machine, so other customers never affect you and responses stay fast and consistent. A shared endpoint splits finite resources across many users, so as traffic rises requests get processed sequentially and latency climbs. Dedicated Bare Metal also unlocks plain non-TLS HTTP, which skips the roughly 20ms TLS handshake that HTTPS pays on every connection, so requests start moving sooner on every call to the leader.

ERPC operates nodes in nine regions: Frankfurt (FRA), Amsterdam (AMS), London (LON), Stockholm (STO), New York (NY), Chicago (CHI), Salt Lake City (SLC), Tokyo (TYO), and Singapore (SGP). Solana leaders rotate across validators around the world, so validator-dense regions such as Frankfurt and the US East Coast see leader slots more often. Use the ERPC Leader Slot API and real measurements to decide between single-region, dual-region, or global deployment for the closest proximity to upcoming leaders.

The lowest-latency setup pairs a dedicated Direct Shreds node with ERPC Bare Metal on the same network for zero-distance communication, with latency as low as approximately 0.1ms ping. Where you place your infrastructure is a decision you revisit rather than make once, so track the leader schedule for your region with the ERPC Leader Slot API (getLeaderSlots) to see which validator leads next before the slot arrives. It returns the upcoming leader schedule, stake weight for the current epoch, estimated validator geolocation, and reference ping values. For details, contact the team via the ERPC Web Dashboard at dashboard.erpc.global.

Your client language has a real impact on how fast you can act on stream data: Rust ranks ahead of Go, then TypeScript, then Python, so Rust is recommended for the highest performance. For the data channel itself, the stream speed order is Direct Shreds first, then Geyser gRPC, then WebSocket, with each one sitting a step closer to the raw network stream. Every plan includes latency-measuring tools in TypeScript and Rust, so you can verify this in your own environment.