AMD EPYC 922424 cores · 128 GB DDR5 ECC · 2 × 1.92 TB NVMe
The balanced node box: 24 Genoa cores, 128 GB and 3.84 TB of NVMe in RAID 0 for chain data — Bitcoin, Monero and an Ethereum pair side by side.
Blockchain nodes
Blockchain node hosting. Full nodes that never fall behind.
Bitcoin, Monero, Ethereum and Solana nodes, validators and RPC endpoints need what a VPS rarely gives: fast NVMe that never throttles, real memory for caches and a port that is not metered. Pay for the machine in the coin you are running.
- NVMedatacenter drives, no IOPS cap
- UnmeteredP2P traffic never counted
- ECCmemory on every server
- $125from, per month
What this workload needsWorkload profile
- Storage that keeps upInitial sync is a random-I/O marathon. Datacenter NVMe finishes it in hours instead of days, and keeps up with new blocks without lag.
- Room for chain dataA Bitcoin node passes 750 GB in 2026 and a comfortable Ethereum execution + consensus pair wants 2 TB of NVMe. Plan for growth, not today’s size.
- Memory for cachesA larger database cache (Bitcoin’s dbcache, Geth’s cache) shortens sync and reduces disk writes. 128 GB leaves plenty for several chains at once.
- Peers and bandwidthNodes serve blocks to hundreds of peers. An unmetered port means no bill for being a good network citizen.
Recommended
Best dedicated servers for blockchain nodes
Chosen by the engineers who size these workloads every week. Every one of them is in stock.
AMD EPYC 9555P64 cores · 256 GB DDR5 ECC · 2 × 3.84 TB NVMe
Archive nodes and multi-chain RPC: 64 Turin cores, 256 GB and 7.68 TB of NVMe.
AMD Ryzen 9 9950X16 cores · 128 GB DDR5 ECC · 2 × 1.92 TB NVMe
High-clock single chain: fastest block validation per core, ideal for a Bitcoin + Lightning or Monero setup.
Advice from our engineers
How to set up a server for blockchain nodes
The first hour on a new server decides how the next year goes.
Where to host it
Frankfurt and Amsterdam are dense in peers and exchanges; Singapore and New York keep latency low for Asian and American users of an RPC endpoint.
Compare latency from your users- Put chain data on RAID 0 (or no RAID): it can always be re-synced, and you double the space. Keep the OS on a small mirrored partition.
- Size caches before the first sync, e.g.
dbcache=8192for Bitcoin Core, then lower it once synced. - Open the P2P ports you need — 8333 (Bitcoin), 18080 (Monero), 30303 (Ethereum execution), 9000 (most consensus clients) — our DDoS filtering leaves legitimate peer traffic untouched.
- Run RPC behind a firewall or reverse proxy: never expose an unauthenticated RPC port to the internet.
FAQ
Questions about blockchain nodes
Still have a question?
Customers ask the engineers who run the servers, in a ticket from the client area — median first reply under 15 minutes, 24/7.
How support worksWhat hardware does a Bitcoin full node need?
Fast SSD storage with room to grow — the chain passes 750 GB in 2026, so plan for 1 TB or more of NVMe — at least 8 GB of RAM (more lets you raise dbcache and sync faster) and an unmetered connection, since a listening node uploads hundreds of gigabytes a month to its peers. Every server we sell covers this; the EPYC 9224 runs Bitcoin, Monero and an Ethereum pair side by side.
Can I pay for a node server with the coin it runs?
Yes. We accept 14 coins, including BTC, XMR, ETH and USDT, with no surcharge.
Can I run a Solana validator or RPC node?
Yes, on the right machine. Solana asks far more of the hardware than Bitcoin or Monero: high-clock cores, hundreds of gigabytes of RAM and separate NVMe drives for accounts and ledger. Look at the EPYC 9455P and 9555P — Turin cores from 3.15 GHz base, memory upgradable to 768 GB, and a 10 Gbps port available in Amsterdam, Frankfurt and New York. Check Anza’s current validator requirements before you order, or ask us for a sizing.
Is running a validator allowed?
Yes. Validators, full nodes, archive nodes, indexers and RPC endpoints are all welcome — it is a whole machine, configured the way your client needs.
What if a disk fails during a sync?
We replace a failed disk within 2 hours. With chain data on RAID 0 you re-sync from peers; the OS on its mirror is unaffected.