NoLimitNodes
PricingDocsBlogAbout
SupportContact
Log in
  1. Home/
  2. Products/
  3. Infrastructure
Compute Platform · VPS / VDS / MetalFrankfurt

Three ways to own Solana compute

Fourteen server shapes, three isolation models, one decision: how much of the silicon do you need to own? Shared vCPU from $100, pinned cores with a dedicated private VLAN to the NLN fleet from $245, the whole machine from $253. This page exists to make that decision quickly and honestly.

Compare the catalogTalk to an engineer
vds.performance · first measurement on the private VLAN
$ ping -c 5 rpc.nln.clr3.org
rtt min/avg/max/mdev = 0.18/0.21/0.27/0.03 ms
$ ping -c 5 api.mainnet-beta.solana.com
rtt avg 14.9 ms · public internet
$ mpstat -P ALL 1 5 | grep -i steal
%steal: 0.00 · every core, every sample
# pinned cores + private VLAN: what VDS and
# bare metal add on top of a VPS.
Catalog
14 SKUs
$100/mo VPS → $12,758/mo H200 GPU
NLN fleet path
Private VLAN
Included on every tier
Provisioning
3 min – 24 h
VDS minutes · VPS same-day · metal next-day
Uptime SLO
99.99%
automatic credits, no ticket required
CompareDecideConfigureNetworkOperationsFAQ
from $100/mo
Three isolation models

Own a slice, own the cores, or own the machine

The families differ in how much silicon is exclusively yours: a shared slice, pinned cores, or the whole machine. Every tier includes a dedicated private VLAN to the NLN fleet. Dashboard, billing, and SLA culture are identical by design.
Shared Compute
VPS
from $100/mo
Deterministic Compute
VDS
from $245/mo
Dedicated Metal
Bare Metal
from $458/mo
TenancyShared host, fair-share vCPUPinned physical coresEntire physical machine
CPU steal timeReal, measure with mpstat0.00%, contractualNo hypervisor exists
RAM ceiling64 GB128 GB · NUMA-local512 GB
Storage1 NVMe volume2–4 NVMe namespaces2 physical NVMe drives
Network1 Gbps · 2 TB egress included5–10 Gbps committed QoS10 Gbps unshared
Path to NLN fleetDedicated private VLANDedicated private VLANDedicated private VLAN
ProvisioningSame-day, most < 1 h2–3 minutes~24 hours, burn-in tested
SLO99.99% monthly99.99% monthly99.99% network & power
Failure recoverySnapshot restore < 15 minLive migration < 8 minIPMI + MTTR < 4 h
Mainnet validatorNoNo (RAM + I/O floor)Yes, built for it
Choose it whenThe workload waits on the networkConsistent compute has a PnLYou need the hardware itself
Explore VPS →Explore VDS →Explore Bare Metal →
the same host, drawn three ways
CPU steal time
0.00%, contractual
Cache & memory bus
your cores, your NUMA node
Failure blast radius
host fails → live migration < 8 min
You pay for
determinism: consistent tail performance
yours time-sliced other tenants host reserve
Sizing, consultatively

Start from the workload, not the price list

Nine real workload profiles, each mapped to the SKU our engineers would put it on, with the reasoning and the honest runner-up.
What are you running?

Bursts CPU on signals, idles between slots. Network-bound 95% of the time.

Latency to RPC/gRPC dominates PnLCPU mostly idleStateless or near-stateless
Our recommendation
vps.pro8 vCPU · 32 GB · 320 GB NVMe
$190/mo

The bot spends its life waiting on the wire. 8 shared vCPU covers the duty cycle with room to spare, and the money stays in the strategy.

Sanity-check with an engineer

Step up ifmpstat shows steal > 1% in peak hours, or your loop decodes every transaction.

vds.standard · $245/mo

Hot loop on the gRPC feed, decode everything, race a single thread to the trigger.

Consistent compute has a direct PnLSustained CPUSingle-thread speed matters
Our recommendation
vds.performance16 cores · 64 GB · 2 × 480 GB NVMe
$495/mo

Pinned cores keep the decode loop consistent: a tail event during your neighbor's burst is a missed trade.

Sanity-check with an engineer

Step up ifthe race comes down to one thread: 4.3 GHz Ryzen metal beats 3.0 GHz virtualized.

nln.metal.turbo · $647/mo

Constant ingestion from Yellowstone into Postgres on the same box. Never idles.

Sustained CPU + diskDB and ingest fight for resourcesRAM for working set
Our recommendation
vds.performance16 cores · 64 GB · 2 × 480 GB NVMe
$495/mo

16 pinned cores and NUMA-local RAM let ingest and Postgres share a box without sharing tails. Unix-socket DB writes, zero network hop.

Sanity-check with an engineer

Step up ifyou index more than a couple of programs or keep > 40 GB hot in Postgres.

vds.max · $895/mo

Receive events, transform, fan out over HTTP. Spiky concurrency, light CPU.

Egress volumeConnection concurrencyCost sensitivity
Our recommendation
vps.enterprise16 vCPU · 64 GB · 640 GB NVMe
$350/mo

16 vCPU absorbs concurrency spikes; shared vCPU is irrelevant when the work is mostly I/O.

Sanity-check with an engineer

Step up ifconcurrency is modest, same architecture, $160 less.

vps.pro · $190/mo

Your own read-only RPC node, ledger and accounts on separate volumes.

128 GB RAM floorSustained disk IOPSCatch-up bandwidth
Our recommendation
vds.max32 cores · 128 GB · 4 × 480 GB NVMe
$895/mo

32 pinned cores, 128 GB, and four NVMe namespaces handle a non-voting mirror, virtualized, so you can resize as it grows.

Sanity-check with an engineer

Step up ifthe mirror is permanent: a full 64-core physical box, physical drives, no hypervisor.

nln.metal.base · $1,227/mo

Learning the operator workflow: vote, restart, snapshot, monitor.

Modest spec is fineCheap to keep runningSame tooling as mainnet
Our recommendation
vps.enterprise16 vCPU · 64 GB · 640 GB NVMe
$350/mo

16 vCPU / 64 GB carries a testnet validator. The point is the runbook practice, not the hardware.

Sanity-check with an engineer

Step up ifyou want the exact I/O behavior you will see on mainnet.

nln.metal.base · $1,227/mo

Voting on mainnet. Skipped slots cost stake. Hypervisor jitter is disqualifying.

256+ GB RAM field standardTwo+ physical NVMeUnshared 10 Gbps
Our recommendation
nln.metal.max64 cores · 256 GB · 2 × 3.84 TB NVMe
$1,902/mo

64 high-frequency EPYC 9575F cores, 256 GB, and dual 3.84 TB NVMe is the field-standard build. The vote pipeline never crosses a hypervisor.

Sanity-check with an engineer

Step up ifyou prefer a dual-socket 64-core layout with the same 256 GB.

nln.metal.dual · $926/mo

Custom Geyser plugin feeding your own downstream: full-chain firehose in, structured data out.

Heavy sustained decodeLarge RAM working setFast local sink
Our recommendation
nln.metal.base64 cores · 128 GB · 2 × 1.92 TB NVMe
$1,227/mo

Full-chain Geyser is a sustained-everything workload: real drives for the sink, real cores for the decode, 128 GB to breathe.

Sanity-check with an engineer

Step up ifyou want more RAM headroom for the working set: 192 GB on the EPYC 9555P.

nln.metal.pro · $1,622/mo

The next validator client: tile architecture, AF_XDP, large memory appetite.

High core countLarge RAM working setKernel-level tuning
Our recommendation
nln.metal.max64 cores · 256 GB · 2 × 3.84 TB NVMe
$1,902/mo

Built for it: 64 high-frequency cores, 256 GB, and a Firedancer-tuned image (governor, IRQ affinity, AF_XDP) on request. Need more RAM? We quote custom builds.

Sanity-check with an engineer
Full catalog · 12 SKUs

Configure across the whole catalog

Every spec and price below is exactly what hits checkout, generated from the same table that does the billing.
Need a shape that isn't on the rail: more drives, GPUs, 100 Gbps? We quote custom builds in 24 h.
vps.starter

The smallest box that runs a real Solana workload.

$100/mo
monthly · cancel anytime
Compute
4 vCPU · shared EPYC
4 vCPU · fair-share scheduled on shared EPYC
Memory
16 GB DDR4 ECC
dedicated allocation, shared channels
Storage
160 GB NVMe
▮ 160 GB NVMe
local NVMe (host mdraid1) · nightly snapshots · restore < 15 min
Network
1 Gbps fair-share2 TB included egress
Contract
  • ·Own kernel, own IP, full root
  • ·Fair-share vCPU and network scheduling
  • ·Local NVMe storage, nightly snapshots
  • ·Key-only SSH from first boot, any OS
  • ·Provisioning: Same-day · most within the hour (EU business hours)
  • ·SLO: 99.99% monthly · automatic credits
Sized for
Testnet botsWebhook handlersMonitoring agentsBuild runners

Not for: Sustained CPU above ~60%; move to VDS before tail latency bites.

Talk to an engineerFrankfurt · root SSH on handover · no setup fee
vps.pro

The production default for trading bots that wait on the network.

$190/mo
monthly · cancel anytime
Compute
8 vCPU · shared EPYC
8 vCPU · fair-share scheduled on shared EPYC
Memory
32 GB DDR4 ECC
dedicated allocation, shared channels
Storage
320 GB NVMe
▮ 320 GB NVMe
local NVMe (host mdraid1) · nightly snapshots · restore < 15 min
Network
1 Gbps fair-share2 TB included egress
Contract
  • ·Own kernel, own IP, full root
  • ·Fair-share vCPU and network scheduling
  • ·Local NVMe storage, nightly snapshots
  • ·Key-only SSH from first boot, any OS
  • ·Provisioning: Same-day · most within the hour (EU business hours)
  • ·SLO: 99.99% monthly · automatic credits
Sized for
Production trading botsLight indexersSniper bots (bursty)

Not for: Hot decode loops at full gRPC volume: that is vds.standard territory.

Talk to an engineerFrankfurt · root SSH on handover · no setup fee
vps.enterprise

Headroom for fanout, simulation, and testnet validator work.

$350/mo
monthly · cancel anytime
Compute
16 vCPU · shared EPYC
16 vCPU · fair-share scheduled on shared EPYC
Memory
64 GB DDR4 ECC
dedicated allocation, shared channels
Storage
640 GB NVMe
▮ 640 GB NVMe
local NVMe (host mdraid1) · nightly snapshots · restore < 15 min
Network
1 Gbps fair-share2 TB included egress
Contract
  • ·Own kernel, own IP, full root
  • ·Fair-share vCPU and network scheduling
  • ·Local NVMe storage, nightly snapshots
  • ·Key-only SSH from first boot, any OS
  • ·Provisioning: Same-day · most within the hour (EU business hours)
  • ·SLO: 99.99% monthly · automatic credits
Sized for
Webhook fanoutMEV simulatorsTestnet validatorsIndexers

Not for: Workloads where consistent tail performance has a direct PnL.

Cross-shopFor $105 less, vds.standard trades 8 shared vCPU for 8 pinned cores.

Talk to an engineerFrankfurt · root SSH on handover · no setup fee
vds.standard

Eight cores nobody else can touch. Steal time: 0.00% by contract.

$245/mo
monthly · cancel anytime
Compute
AMD EPYC · 8 cores @ 3.0 GHz · pinned
8 physical cores · pinned via cgroups · single NUMA node
Memory
32 GB DDR4 ECC
allocated NUMA-local to your cores
Storage
2 × 250 GB NVMe
▮ 250 GB NVMe▮ 250 GB NVMe
dedicated NVMe namespaces, local to the host
Network
5 Gbps committed10 TB included egressprivate VLAN to the NLN fleet
Contract
  • ·Physical cores pinned, single socket, no shared hyperthreads
  • ·RAM allocated from the same NUMA node as your cores
  • ·Committed bandwidth in the QoS class, not borrowed
  • ·Dedicated NVMe namespaces, verifiable with lscpu / numactl / mpstat
  • ·Provisioning: 2–3 minutes · capacity is held hot
  • ·SLO: 99.99% monthly · host-failure migration target < 8 min
Sized for
Hot trading loopsFull-volume gRPC consumersLatency-sensitive single services

Not for: Bursty, mostly-idle workloads: a vps.pro does that for $55 less.

Talk to an engineerFrankfurt · root SSH on handover · no setup fee
vds.performance

The indexer workhorse: ingest, decode, and Postgres on one NUMA node.

$495/mo
monthly · cancel anytime
Compute
AMD EPYC · 16 cores @ 3.0 GHz · pinned
16 physical cores · pinned via cgroups · single NUMA node
Memory
64 GB DDR4 ECC
allocated NUMA-local to your cores
Storage
2 × 480 GB NVMe
▮ 480 GB NVMe▮ 480 GB NVMe
dedicated NVMe namespaces, local to the host
Network
10 Gbps committedUnmetered egressprivate VLAN to the NLN fleet
Contract
  • ·Physical cores pinned, single socket, no shared hyperthreads
  • ·RAM allocated from the same NUMA node as your cores
  • ·Committed bandwidth in the QoS class, not borrowed
  • ·Dedicated NVMe namespaces, verifiable with lscpu / numactl / mpstat
  • ·Provisioning: 2–3 minutes · capacity is held hot
  • ·SLO: 99.99% monthly · host-failure migration target < 8 min
Sized for
Indexer + Postgres co-locatedMEV searchersGeyser consumers (1–2 programs)

Not for: Mainnet validators. The RAM ceiling and shared drives rule it out.

Talk to an engineerFrankfurt · root SSH on handover · no setup fee
vds.max

The top of virtualization. Past this point, buy the machine.

$895/mo
monthly · cancel anytime
Compute
AMD EPYC · 32 cores @ 3.0 GHz · pinned
32 physical cores · pinned via cgroups · single NUMA node
Memory
128 GB DDR4 ECC
allocated NUMA-local to your cores
Storage
4 × 480 GB NVMe
▮ 480 GB NVMe▮ 480 GB NVMe▮ 480 GB NVMe▮ 480 GB NVMe
dedicated NVMe namespaces, local to the host
Network
10 Gbps committedUnmetered egressprivate VLAN to the NLN fleet
Contract
  • ·Physical cores pinned, single socket, no shared hyperthreads
  • ·RAM allocated from the same NUMA node as your cores
  • ·Committed bandwidth in the QoS class, not borrowed
  • ·Dedicated NVMe namespaces, verifiable with lscpu / numactl / mpstat
  • ·Provisioning: 2–3 minutes · capacity is held hot
  • ·SLO: 99.99% monthly · host-failure migration target < 8 min
Sized for
RPC mirrors (non-voting)Full-chain Geyser pipelinesHeavy multi-service stacks

Not for: Mainnet voting or >128 GB RAM: step up to a physical 64-core machine.

Cross-shopnln.metal.base: a full 64-core EPYC 9554P, 128 GB, physical NVMe, no hypervisor at all.

Talk to an engineerFrankfurt · root SSH on handover · no setup fee
nln.metal.edge

Entry metal: high clocks, your own NIC, no hypervisor anywhere.

$253/mo
monthly · cancel anytime
Compute
Intel Xeon E-2286G · 6 cores @ 4.0 GHz
6 physical cores · entire machine · no hypervisor
Memory
32 GB DDR4 ECC
every channel is yours
Storage
2 × 960 GB SSD
▮ 960 GB NVMe▮ 960 GB NVMe
physical drives; workloads never share IOPS across bays
Network
10 Gbps committedUnmetered egressprivate VLAN to the NLN fleet
Contract
  • ·Full control of IRQ affinity, CPU governor, and kernel
  • ·Physical NVMe drives; ledger and accounts DB never share IOPS
  • ·10–100 Gbps unshared host networking
  • ·IPMI / KVM-over-IP on a dedicated management VLAN
  • ·Provisioning: ~24 hours · OS installed, drives validated, burn-in tested
  • ·SLO: 99.99% network & power · hardware MTTR < 4 h (business window)
Sized for
Latency-critical single servicesDedicated relaysSmall Geyser pipelines

Not for: Core-hungry parallel decode: six cores is six cores.

Talk to an engineerFrankfurt · root SSH on handover · no setup fee
nln.metal.core

A steady multi-core box for always-on services.

$346/mo
monthly · cancel anytime
Compute
Intel Xeon Silver 4214 · 12 cores @ 2.2 GHz
12 physical cores · entire machine · no hypervisor
Memory
64 GB DDR4 ECC
every channel is yours
Storage
2 × 960 GB SSD
▮ 960 GB NVMe▮ 960 GB NVMe
physical drives; workloads never share IOPS across bays
Network
10 Gbps committedUnmetered egressprivate VLAN to the NLN fleet
Contract
  • ·Full control of IRQ affinity, CPU governor, and kernel
  • ·Physical NVMe drives; ledger and accounts DB never share IOPS
  • ·10–100 Gbps unshared host networking
  • ·IPMI / KVM-over-IP on a dedicated management VLAN
  • ·Provisioning: ~24 hours · OS installed, drives validated, burn-in tested
  • ·SLO: 99.99% network & power · hardware MTTR < 4 h (business window)
Sized for
IndexersWebhook fanoutBackground workers

Not for: Single-thread race paths: clocks are modest here.

Talk to an engineerFrankfurt · root SSH on handover · no setup fee
nln.metal.turbo

The highest clocks in the fleet. Built for single-thread race paths.

$647/mo
monthly · cancel anytime
Compute
AMD Ryzen 9 9950X · 16 cores @ 4.3 GHz
16 physical cores · entire machine · no hypervisor
Memory
192 GB DDR4 ECC
every channel is yours
Storage
2 × 1.92 TB NVMe
▮ 1.92 TB NVMe▮ 1.92 TB NVMe
physical drives; workloads never share IOPS across bays
Network
10 Gbps committedUnmetered egressprivate VLAN to the NLN fleet
Contract
  • ·Full control of IRQ affinity, CPU governor, and kernel
  • ·Physical NVMe drives; ledger and accounts DB never share IOPS
  • ·10–100 Gbps unshared host networking
  • ·IPMI / KVM-over-IP on a dedicated management VLAN
  • ·Provisioning: ~24 hours · OS installed, drives validated, burn-in tested
  • ·SLO: 99.99% network & power · hardware MTTR < 4 h (business window)
Sized for
MEV hot pathsSniper enginesAnything where one thread decides the trade

Not for: Wide 64-core parallel decode: pick an EPYC tier for that.

Talk to an engineerFrankfurt · root SSH on handover · no setup fee
nln.metal.dual

64 cores across two sockets with 256 GB for wide parallel work.

$926/mo
monthly · cancel anytime
Compute
Dual AMD EPYC 7452 · 64 cores @ 2.35 GHz
64 physical cores · entire machine · no hypervisor
Memory
256 GB DDR4 ECC
every channel is yours
Storage
2 × 3.84 TB NVMe
▮ 3.84 TB NVMe▮ 3.84 TB NVMe
physical drives; workloads never share IOPS across bays
Network
10 Gbps committedUnmetered egressprivate VLAN to the NLN fleet
Contract
  • ·Full control of IRQ affinity, CPU governor, and kernel
  • ·Physical NVMe drives; ledger and accounts DB never share IOPS
  • ·10–100 Gbps unshared host networking
  • ·IPMI / KVM-over-IP on a dedicated management VLAN
  • ·Provisioning: ~24 hours · OS installed, drives validated, burn-in tested
  • ·SLO: 99.99% network & power · hardware MTTR < 4 h (business window)
Sized for
Full-chain Geyser pipelinesIndexers + PostgresNon-voting RPC mirrors

Not for: Latency-bound single threads: 2.35 GHz is the trade-off for core count.

Talk to an engineerFrankfurt · root SSH on handover · no setup fee
nln.metal.base

64 modern EPYC cores: the RPC / validator baseline.

$1,227/mo
monthly · cancel anytime
Compute
AMD EPYC 9554P · 64 cores @ 3.1 GHz
64 physical cores · entire machine · no hypervisor
Memory
128 GB DDR4 ECC
every channel is yours
Storage
2 × 1.92 TB NVMe
▮ 1.92 TB NVMe▮ 1.92 TB NVMe
physical drives; workloads never share IOPS across bays
Network
10 Gbps committedUnmetered egressprivate VLAN to the NLN fleet
Contract
  • ·Full control of IRQ affinity, CPU governor, and kernel
  • ·Physical NVMe drives; ledger and accounts DB never share IOPS
  • ·10–100 Gbps unshared host networking
  • ·IPMI / KVM-over-IP on a dedicated management VLAN
  • ·Provisioning: ~24 hours · OS installed, drives validated, burn-in tested
  • ·SLO: 99.99% network & power · hardware MTTR < 4 h (business window)
Sized for
RPC mirrorsGeyser + Postgres pipelinesTestnet validators

Not for: Mainnet voting; 128 GB is below the 256 GB field standard.

Talk to an engineerFrankfurt · root SSH on handover · no setup fee
nln.metal.pro

64 cores with more headroom for heavy RPC and indexer stacks.

$1,622/mo
monthly · cancel anytime
Compute
AMD EPYC 9555P · 64 cores @ 3.2 GHz
64 physical cores · entire machine · no hypervisor
Memory
192 GB DDR4 ECC
every channel is yours
Storage
2 × 1.92 TB NVMe
▮ 1.92 TB NVMe▮ 1.92 TB NVMe
physical drives; workloads never share IOPS across bays
Network
10 Gbps committedUnmetered egressprivate VLAN to the NLN fleet
Contract
  • ·Full control of IRQ affinity, CPU governor, and kernel
  • ·Physical NVMe drives; ledger and accounts DB never share IOPS
  • ·10–100 Gbps unshared host networking
  • ·IPMI / KVM-over-IP on a dedicated management VLAN
  • ·Provisioning: ~24 hours · OS installed, drives validated, burn-in tested
  • ·SLO: 99.99% network & power · hardware MTTR < 4 h (business window)
Sized for
Full RPC nodesHeavy indexer pipelinesMulti-service stacks
Talk to an engineerFrankfurt · root SSH on handover · no setup fee
nln.metal.max

High-frequency 64-core, 256 GB: the mainnet validator turnkey.

$1,902/mo
monthly · cancel anytime
Compute
AMD EPYC 9575F · 64 cores @ 3.3 GHz
64 physical cores · entire machine · no hypervisor
Memory
256 GB DDR4 ECC
every channel is yours
Storage
2 × 3.84 TB NVMe
▮ 3.84 TB NVMe▮ 3.84 TB NVMe
physical drives; workloads never share IOPS across bays
Network
10 Gbps committedUnmetered egressprivate VLAN to the NLN fleet
Contract
  • ·Full control of IRQ affinity, CPU governor, and kernel
  • ·Physical NVMe drives; ledger and accounts DB never share IOPS
  • ·10–100 Gbps unshared host networking
  • ·IPMI / KVM-over-IP on a dedicated management VLAN
  • ·Provisioning: ~24 hours · OS installed, drives validated, burn-in tested
  • ·SLO: 99.99% network & power · hardware MTTR < 4 h (business window)
Sized for
Mainnet validators (Agave)FiredancerFull RPC nodes
Talk to an engineerFrankfurt · root SSH on handover · no setup fee
nln.metal.gpu

NVIDIA H200 on dual 32-core EPYC, 512 GB: GPU compute.

$12,758/mo
monthly · cancel anytime
Compute
NVIDIA H200 · 2 × EPYC 32 cores @ 3.55 GHz
64 physical cores · entire machine · no hypervisor
Memory
512 GB DDR4 ECC
every channel is yours
Storage
2 × 1.92 TB NVMe
▮ 1.92 TB NVMe▮ 1.92 TB NVMe
physical drives; workloads never share IOPS across bays
Network
10 Gbps committedUnmetered egressprivate VLAN to the NLN fleet
Contract
  • ·Full control of IRQ affinity, CPU governor, and kernel
  • ·Physical NVMe drives; ledger and accounts DB never share IOPS
  • ·10–100 Gbps unshared host networking
  • ·IPMI / KVM-over-IP on a dedicated management VLAN
  • ·Provisioning: ~24 hours · OS installed, drives validated, burn-in tested
  • ·SLO: 99.99% network & power · hardware MTTR < 4 h (business window)
Sized for
DePIN / GPU workloadsOn-chain ML & inferenceAccelerated analytics
Talk to an engineerFrankfurt · root SSH on handover · no setup fee
Every tier

What the private VLAN buys you

Every NLN server — VPS, VDS, and bare metal — sits on the same switch fabric as the NLN validator and RPC fleet, with a dedicated private VLAN leg. This is that room, and what the wire costs from inside it.
network topology · frankfurt● all paths measured
EQUINIX-CLASS FACILITY · FRANKFURTone room. one switch fabric. everything below is meters apart.Your VDS / machineroot SSH · private VLAN leg10.x.x.x → switch fabricVPS · VDS · BARE METALNLN RPC · gRPC · WS fleetrpc / grpc / ws .nln.clr3.orgthe endpoints your code callsNLN validator nodeswhere slots are actually heardfirst. same switch fabric.Jito block engine0.6 ms · cross-connectprivate VLANsame fabric · sub-mspublic internetapi.mainnet-beta.solana.com · 14.2 msleaves the building · 14+ ms · long-tail jitter
NLN RPC / gRPC fleetPrivate VLAN
burns NaN% of your ~100 ms reaction window per round trip
NLN validator nodesSame switch fabric
burns NaN% of your ~100 ms reaction window per round trip
Jito block engine (FRA)Cross-connect
burns NaN% of your ~100 ms reaction window per round trip
api.mainnet-beta.solana.comPublic internet
burns NaN% of your ~100 ms reaction window per round trip
log scale · medians · slot time on Solana: 400 ms · last verified 2026-05-12
inside the building, the wire is effectively free
Operational maturity

One SLA culture across the fleet

Capacity policy, hardware lifecycle, failure recovery: written down, measured, and identical whether you rent four vCPUs or a 512 GB machine.
§Capacity policy
  • ·VPS pools run fair-share vCPU scheduling with headroom held in reserve
  • ·VDS and metal capacity is never shared; pinned capacity is held hot
  • ·Spare capacity stays racked and powered, so orders never wait for a host to be carved
§Hardware lifecycle
  • ·AMD EPYC platforms, DDR4 ECC across the fleet
  • ·Datacenter-grade storage with redundancy (mdraid1) on virtualized pools
  • ·Every metal box burn-in tested before handover: CPU, RAM, drive surface, NIC
§Failure & recovery
  • ·VPS: nightly snapshots, restore to a healthy host < 15 min
  • ·VDS: live migration off a failing host, target < 8 min
  • ·Metal: IPMI / KVM-over-IP always on, hardware MTTR < 4 h in the business window
§Network operations
  • ·Committed QoS bandwidth on VDS; 10–100 Gbps dedicated ports on metal
  • ·Dedicated private VLAN to the NLN RPC / gRPC / WS fleet included on every VDS and bare-metal server
  • ·Fair-share scheduler on shared uplinks; one tenant cannot saturate a host
§Access & control
  • ·Full root. No locked kernels, no restricted outbound ports, no inspection
  • ·Key-only SSH from first boot; password auth ships disabled
  • ·Bring your own ISO: Ubuntu, Debian, Rocky, Arch, NixOS all routine
§SLO & credits
  • ·99.99% monthly rolling, fleet-wide
  • ·Credits applied automatically on breach, no ticket required
  • ·Only enforcement on your box: outbound SMTP abuse limits

Frequently asked questions

By workload, not budget. If the process spends most of its life waiting on network events (trading bots, webhooks, monitoring), shared vCPU is invisible and a VPS from $100/mo is right. If the process runs hot and its 99th-percentile latency has a cost (snipers, full-volume gRPC consumers, indexers), you want pinned cores and contractual 0.00% steal time, which is VDS from $245/mo. If the workload is a mainnet validator, Firedancer, or anything needing 128+ GB RAM and physical drives, take the whole machine: bare metal from $253/mo. The decision engine on this page walks the same logic interactively.
Because that is where our validator and RPC fleet lives. Every tier ships with a dedicated private VLAN leg that reaches rpc.nln.clr3.org and grpc.nln.clr3.org without leaving the switch fabric, and from those racks, the Jito block engine is a cross-connect in the same building. VPS, VDS, and bare metal are all validator-adjacent in Frankfurt. We can quote bespoke regions for committed annual contracts.
Monthly subscriptions through the dashboard, card via Stripe or crypto. The prices on these pages are the prices at checkout: the catalog is generated from the same table that does the billing, so there are no configurator surprises. Cancel anytime; SLO credits apply automatically on breach without a ticket.
Yes, and it is designed to be boring. A VPS-to-VDS migration is an rsync and an endpoint swap: 15 to 30 minutes. VDS resizes take minutes because dedicated capacity is held hot. Moving onto metal is a fresh provision (~24 hours) plus the same data copy.
Full root over SSH with key-only auth from first boot, your choice of OS (Ubuntu, Debian, Rocky, or your own ISO), DDoS-filtered transit, and an SLA with automatic credits. Every tier includes a dedicated private VLAN to the NLN RPC/gRPC/WebSocket fleet at no extra cost. We don't lock kernels, restrict outbound ports, or inspect workloads; the only enforcement is outbound SMTP abuse limits.
Yes. GPUs (H100, L40S), 1 TB RAM builds, 100 Gbps fabric, multi-rack deployments: anything available from Supermicro, Dell, or HPE racks within 7–10 days, quoted by an engineer within 24 hours, and lands on the same private VLAN as the catalog SKUs.

Related products

Solana VPS

Shared compute from $100/mo for bots, webhooks, and indexers.

Solana VDS

Deterministic compute from $245/mo. Pinned cores, 0.00% steal time.

Bare-metal servers

The whole machine from $253/mo. Validator-, Firedancer-, and GPU-grade.

Solana RPC nodes

The endpoint every NLN server reaches over the private VLAN.

Yellowstone gRPC nodes

The streaming layer trade-adjacent workloads consume.

Geyser plugin hosting

Managed plugin slots on production validators.

Pick a tier and deploy today

VPS from $100/mo, VDS from $245/mo, bare metal from $253/mo. One dashboard, one bill, one support queue. VDS provisions in minutes; most VPS orders ship within the hour.

Talk to an engineer

Not sure between two SKUs? Send the workload description and an engineer replies with a sizing call, usually same day.

order → root ssh
  1. 01
    Create an account
    Email and a password. No card required to look around.
  2. 02
    Pick tier + paste your SSH key
    The exact specs and prices on this page. No checkout surprises.
  3. 03
    Root SSH lands in your inbox
    Key-only auth, your chosen OS image, ready for the NLN data fleet.

Ready to get started?

Get your free API key and start building in under 30 seconds.

Talk to Sales
NoLimitNodes

Solana RPC infrastructure built for performance and scale.

RPC Access
  • HTTP RPC
  • WebSocket
  • gRPC
Infrastructure
  • Compute Platform
  • VPS
  • VDS
  • Bare Metal
  • Geyser Plugin Hosting
Enhanced Streams
  • PumpFun
  • PumpSwap
  • Raydium
  • Orca
  • Meteora
  • System Events
  • Browse All →
Program Streams
  • PumpFun
  • PumpSwap
  • Raydium CLMM
  • Orca Whirlpool
  • Meteora DLMM
  • Jupiter Swap
  • Jupiter Perps
  • Kamino Lending
  • Browse All 37 →
Trading
  • EZWallet
Analytics
  • Historical Datasets
  • Historical Raw Blocks
Company
  • About
Resources
  • Pricing
  • Custom Development
  • Documentation
  • Blog
  • Support
  • Contact Sales
Compare
  • Yellowstone gRPC vs LaserStream
  • Triton vs Helius
  • Raydium API vs Helius
  • PumpSwap API vs Bitquery
  • QuickNode Streams vs NLN
  • All comparisons →
Legal
  • Terms & Conditions
  • Privacy Policy
© 2026 CLR3 Inc., operating as NoLimitNodes. Registered in Ontario, Canada. All rights reserved.solana mainnet