Back to blog
SolanaFebruary 10, 2026·4 min read

Solana Validator Server Requirements: Hardware, Bandwidth & Location

Solana Validator Server Requirements: Hardware, Bandwidth & Location

Running a Solana validator is one of the most demanding things you can ask of a server. The network expects validators to keep pace with rapid block production, vote reliably, and serve data - all continuously, without falling behind. Under-spec the hardware and you will see skipped slots, delinquency, and missed rewards. This guide lays out what you actually need, and why.

CPU: clocks first, then cores

Solana's runtime benefits heavily from strong single-thread performance, so high clock speeds matter more than raw core count. The Agave client's published baseline is 12 cores / 24 threads at a 2.8GHz base clock or faster, on AMD Zen 3 (or newer) or Intel Ice Lake (or newer), with SHA and AVX2 instruction support. In practice, production mainnet operators run well beyond the minimum - 24-core-plus AMD EPYC parts at 3.5GHz and up are the community standard. The guiding principle from the docs is explicit: higher clock speed is preferable to more cores. Our water-cooled EPYC 4565P is a good example of the high-clock, sustained-performance profile validators want.

Memory: plan generously

Validators are memory-hungry. Account state lives in RAM, and running tight on memory leads to instability and performance cliffs. For a mainnet validator, provision 256GB or more of ECC memory and treat that as a floor rather than a target - many production operators run 384-512GB. If you also serve RPC traffic with full account indexes, plan for 512GB or more. ECC is strongly recommended for the reliability a long-running validator needs.

Storage: enterprise NVMe, properly laid out

This is where many validators go wrong. Solana is extraordinarily I/O-intensive, and consumer or SATA drives simply cannot keep up - and they wear out fast under validator write loads. The Agave docs call for PCIe Gen3 x4 NVMe or better, with high TBW (Total Bytes Written) endurance.

  • Use enterprise-grade NVMe rated for high sustained writes.
  • Separate the ledger and accounts onto different physical drives - the high IOPS make sharing one disk a bottleneck and it shortens drive life.
  • Provision at least 1TB each for the accounts and ledger drives (more if you keep longer transaction history), plus space for snapshots; the ledger grows continuously.

Bandwidth and networking

A validator continuously exchanges large volumes of data with the rest of the cluster. The recommended minimum is 1 Gbps symmetric, with 10 Gbps preferred for production, on a dedicated public IP. Just as important as raw capacity is strong peering to minimise packet loss and jitter. Poor networking shows up as missed votes and degraded performance even when your CPU and storage are fine. Premium peering - like the direct Teraswitch connections in our network - keeps paths short and stable.

Location

Choose a location with excellent connectivity to the rest of the validator set. Network position affects how quickly you receive and propagate data across the cluster. Our high-memory, premium-peered AMD EPYC bare metal in Frankfurt and Amsterdam makes a strong validator home; see the location guide to choose. (Our high-clock Ryzen boxes in New York and Salt Lake City are tuned for trading bots and RPC clients rather than full validators, which need 256GB+ of ECC memory.)

Recommended baseline (mainnet)

ComponentRecommended
CPUHigh-clock AMD EPYC, 12+ cores / 24 threads (24+ cores in production), 2.8GHz+ base
RAM256GB+ ECC (384-512GB in production)
StorageEnterprise NVMe (PCIe Gen3 x4+), 1TB+ each on separate ledger + accounts drives, high TBW
Network1 Gbps minimum, 10 Gbps preferred, premium peering, dedicated IP
PlatformDedicated bare metal (not a VPS)

FAQ

Can I run a validator on a VPS?

For mainnet, no - a validator needs the guaranteed CPU, RAM, and I/O of dedicated bare metal. A VPS is fine for testnet experimentation or lightweight RPC clients.

Why does storage matter so much?

Solana's write volume is extreme. Cheap drives bottleneck performance and wear out quickly, which is why enterprise NVMe and a sensible drive layout are essential.

How do I get help sizing a validator?

Open a ticket with us. We build validator-grade bare metal regularly and can tailor a configuration to your needs.

Build it with Orbit Servers

We provide premium-peered, NVMe-backed bare metal ideal for validators across multiple regions. Open a ticket for a custom validator spec, or browse our bare metal options.

Need dedicated bare metal?

Get the whole machine - guaranteed CPU, RAM, and NVMe I/O with premium peering. Ideal for full RPC nodes and Solana validators.

Explore bare metal
O

Written by

Ory

The Orbit Servers Team

The Orbit Servers team builds and operates low-latency VPS, bare metal, and colocation infrastructure across the US, EU, and APAC - with a focus on Solana RPC, validator, and trading workloads.

Related posts