The Kernel
Behind Every Workload

Rent CPU compute by the hour. 2 to 32 vCPU, up to 256 GB RAM, 3 or 5 GHz. Pay with Solana, connect in 60 seconds. No contracts, no KYC.

Own the Compute Layer
of Clusters

The Spark allocates capital across four layers of decentralized AI compute

Compute
Optimized

High-clock cores up to 5 GHz for CPU-bound builds, simulations, and single-threaded jobs. Available by the hour.

General Purpose

Balanced vCPU and memory for web services, CI runners, data pipelines, and everyday backend work.

Memory Optimized

Up to 8 GB RAM per vCPU for in-memory datasets, caches, and memory-heavy jobs. Scale to 256 GB.

Pay With Solana

Pay with Solana — no card, no KYC. Your instance shows up in your dashboard with SSH and Jupyter access.

The Kernel
Lineup

Compute · 32 vCPU

32 vCPU at 5 GHz with 64 GB RAM. Compute-Optimized peak for the heaviest CPU-bound jobs. Billed by the hour.

View Project

Compute · 16 vCPU

16 vCPU at 5 GHz with 32 GB RAM. The everyday workhorse for builds, tests, and services.

View Project

General · 8 vCPU

8 vCPU with 32 GB RAM. Balanced cores and memory, ready in under a minute.

View Project

General · 16 vCPU

16 vCPU with 64 GB RAM. General Purpose throughput for pipelines and parallel workloads.

View Project

Memory · 8 vCPU

8 vCPU with 64 GB RAM. Memory-Optimized for in-memory datasets and caches.

View Project

Starter · 2 vCPU

2 vCPU with 4 GB RAM. The cheapest way in, perfect for dev and batch jobs.

View Project

>

>

Compute, on tap

Real CPU compute rented by the hour. Pay with Solana, connect in 60 seconds. No contracts, no minimums.

60-Second Setup

Pick a config, pay, get a live instance in under a minute. No tickets, no waiting list.

Pay With Solana

Send Solana directly from your wallet. No KYC, no card on file, no monthly minimums.

Pre-Built Images

Start with Ubuntu, Python, or your own Docker image. Ships ready for your workload out of the box.

Hourly Billing

Rent by the hour. Stop the pod when you’re done — billing stops the moment it does.

SSH & Jupyter

Every rental ships with SSH and Jupyter access on a public endpoint. Full root.

Global Regions

Capacity in US East/West, EU, and APAC — pick the region closest to your data.

Real Capacity

Pricing in USD, settled in Solana. What you see at quote time is what you pay — no surge pricing, no hidden fees.

>

>

Kernel CLI

Rent compute from
your terminal

One command. Pick a config, pay in Solana, get SSH back. No browser, no dashboard, no wallet extension. A holder-only tool for people who live in the shell.

Install
curl -fsSL https://kernelcpu.xyz/cli.sh | sh
  • Zero dependencies. One Python file. No Node, no Docker, install in seconds.
  • Pay in Solana. The CLI shows a wallet and an exact amount. You send it, it verifies on-chain, you get the pod.
  • Your key, your pod. Your SSH key is injected at boot — direct root access, no shared credentials.
Holder-only access — requires 1,000,000 KERNEL once the token is live.
$ kernel rent --vcpu 8 --type compute --hours 4 \
    --wallet 7Qf…9xPq

  Kernel · renting compute
  Compute · 3 GHz · 8 vCPU · 16 GB RAM · 4h

  Send exactly
    0.027429833 SOL
  to
    84Zy…UMQUX
  last digits tag your order — send the exact value

  Press Enter once you've sent the payment ▸ 

  Verifying payment on-chain... ✓ found
  Provisioning... 

  ✓ Instance ready
  SSH   ssh root@69.30.85.116 -p 22041