Nova
Nova
On-Demand GPU Compute

Rent a GPU.
Run anything.

Bare-metal GPU power billed by the hour. No signups, no KYC. Connect your Solana wallet, pick hardware, deploy in under a minute. Full root SSH with CUDA pre-installed.

500+
GPUs Available
<60s
Deploy Time
No KYC
Wallet Only
Pay in
SOL
Capabilities

Built for serious compute

Infrastructure for developers and researchers who need GPU power without the cloud overhead.

01

Instant Provisioning

Your GPU goes live in under 60 seconds. No cold boots, no waiting rooms. Select hardware, pay in SOL, connect via SSH.

02

Hourly Billing

Pay per hour with no minimum commitment. Spin up for an hour or a week. Billing stops the moment you terminate.

03

No KYC Required

Your Solana wallet is your identity. No emails, no verification, no credit cards. Connect and deploy.

04

Multi-GPU Options

Scale from a single card to multi-GPU setups. High-end hardware including A100s, H100s, and RTX 4090s.

05

Full Root Access

Bare-metal SSH access with full root privileges. Run whatever you want: training, inference, rendering.

06

Pre-Built Stacks

CUDA, PyTorch, and standard ML libraries pre-installed. Or bring your own Docker image. Zero setup friction.

Hardware

Choose your GPU

Value
RTX 4090
$0.34/hr
Memory24 GB GDDR6X
Compute83 TFLOPS FP32
InterconnectPCIe 4.0 x16
Best ForInference, Rendering
Popular
A100 80GB
$1.10/hr
Memory80 GB HBM2e
Compute19.5 TFLOPS FP32
InterconnectNVLink 600GB/s
Best ForTraining, Fine-tuning
Fastest
H100 SXM
$2.50/hr
Memory80 GB HBM3
Compute67 TFLOPS FP32
InterconnectNVLink 900GB/s
Best ForLarge-scale Training
Process

Zero to running

01

Connect Wallet

Connect your Solana wallet (Phantom, Solflare). No signup, no email, no KYC. Your wallet is your account.

02

Pick Hardware

Browse available GPUs filtered by VRAM, price, and location. Live pricing from community GPU providers.

03

Pay in SOL

One transaction covers a 1-hour deposit. Instance spins up immediately. CUDA pre-installed and ready.

04

SSH and Build

Full root SSH access within 60 seconds. Run training, inference, rendering. Terminate anytime from dashboard.

GPU

We built Nova because renting a GPU shouldn't require a credit card, an account, or a sales call. Connect wallet. Pick hardware. Deploy. Pay in SOL, get SSH access, start computing. That's it.

Every workload,
one platform

Whatever you are building, Nova has the hardware config, the software stack, and the regional availability to run it.

AI Research

Train and fine-tune foundation models with multi-node clusters and high-bandwidth interconnect.

3D Rendering

Render complex scenes and VFX with dedicated GPU power. Deadline farms that spin up in seconds.

Genomics

Accelerate sequence alignment, variant calling, and protein folding with GPU-optimized pipelines.

Autonomous Systems

Develop perception models for self-driving and robotics with massive parallel simulation.

FAQ

Common questions

High-end NVIDIA GPUs including RTX 4090, A100, H100, A6000, and L40. Availability and pricing update in real-time from community GPU providers.

You pay a 1-hour deposit in SOL upfront. The instance runs until you terminate it. We don't auto-charge, you control when it stops.

No. Your Solana wallet is your identity. No email, no password, no KYC. Connect and rent.

Instances come with CUDA and standard ML tools. You get full root SSH access so you can install anything else you need.

Typically under 60 seconds from payment confirmation to SSH access. Some instances may take slightly longer depending on provider.

Instance is destroyed immediately and billing stops. Any data on the instance is lost. Download what you need before terminating.

Phantom and Solflare. Any Solana wallet that supports standard transaction signing should work.

The minimum deposit covers 1 hour. After that you can terminate anytime. No long-term commitments.

GPUs are sourced from community providers globally. You can see the region for each listing on the marketplace page.

Yes. RTX 4090s are great for inference workloads at a lower price point. Pick based on your VRAM and compute needs.

Ready to compute?

Deploy your first GPU in under a minute. Connect wallet, pay in SOL, start computing.