FAQ
About Nirvana Labs
Section titled “About Nirvana Labs”Q: What is Nirvana Labs?
Section titled “Q: What is Nirvana Labs?”A: Nirvana Labs is a high-performance bare-metal cloud purpose-built for Web3. It delivers ultra-low-latency compute, elastic block storage, and private fiber networking for sequencers, validators, routers, and intent execution engines.
Q: What makes Nirvana different from AWS or GCP?
Section titled “Q: What makes Nirvana different from AWS or GCP?”A: Hyperscalers were designed for web workloads — Nirvana is built for real-time, high-throughput blockchain execution.
- Dedicated bare metal (no noisy neighbors)
- Lightweight hypervisor (no virtualization latency)
- Hyper-converged architecture (dedicated compute + storage + networking unified)
- Private routing via Nirvana Connect
- Deterministic performance at lower cost
Q: Where are your data centers?
Section titled “Q: Where are your data centers?”A: Nirvana operates globally with coverage in North America, Europe, and Asia — positioned near major chain hubs for ultra-low-latency connectivity.
We leverage 6 data centers in Silicon Valley, Chicago, Washington DC, Frankfurt, Tokyo, and Singapore.
Q: Who uses Nirvana today?
Section titled “Q: Who uses Nirvana today?”A: Over 50+ teams use Nirvana for their mission-critical blockchain infrastructure — including Aori, BitGo, Chainlink, Fireblocks, Nansen, RedStone, Keyrock, and Goldsky.
These teams run high-performance workloads such as routing engines, intent execution, indexing, validation, and settlement on Nirvana Cloud.
Nirvana also supports 60+ blockchain networks across L1s, L2s, appchains, and testnets — enabling global deployments and elastic scaling for multi-chain infrastructure.
Q: Are you SOC audited?
Section titled “Q: Are you SOC audited?”A: Yes — Nirvana Labs is SOC 2 Type II compliant.
Q: How long does implementation take?
Section titled “Q: How long does implementation take?”A: Standard implementation typically takes 3–4 weeks if all requirements are provided.
Q: What kind of implementation support do you provide?
Section titled “Q: What kind of implementation support do you provide?”A: We provide Slack channel onboarding with engineers across regions to assist you through the entire setup.
Q: How do I get in touch?
Section titled “Q: How do I get in touch?”Ready to Get Started? Get in touch and we’ll help you get set up with Nirvana.
Learn more at
Nirvana | Blog | Docs | Twitter | LinkedIn
Product - General
Section titled “Product - General”Q: What is the Nirvana Cloud stack?
Section titled “Q: What is the Nirvana Cloud stack?”A: Nirvana provides dedicated bare-metal infrastructure with built-in compute, storage, networking, orchestration, observability, and data delivery — designed specifically for blockchain performance and cost efficiency.
A Web3-native cloud stack built for performance
L0 · Bare Metal — Dedicated servers with a lightweight hypervisor, placed near major chain hubs for maximum speed and control.
L1 · Compute + Storage + Network — VMs, Accelerated Block Storage (ABS), and private interconnect (Nirvana Connect) running on a high-performance SDN backbone.
L2 · Orchestration — Fast, automated provisioning with NKS and Terraform.
L3 · Observability — Built-in monitoring, alerting, and failover to keep workloads stable.
L4 · Managed Data Services — Proprietary CDN, API gateway, dedicated nodes and custom node pools for low-latency blockchain data delivery.
Q: What is bare metal cloud?
Section titled “Q: What is bare metal cloud?”A: Bare metal cloud is a cloud computing model that gives you access to dedicated physical servers, not shared or virtualized machines. Unlike traditional cloud (e.g., Amazon Web Services or Google Cloud Platform), where workloads run on a shared hypervisor, bare metal gives you full control of the entire server - CPU, memory, storage, and networking - with no noisy neighbors or virtualization overhead.
Q: What’s special about your bare metal cloud?
Section titled “Q: What’s special about your bare metal cloud?”- No shared tenancy or heavy virtualization
- Low-jitter, deterministic performance
- Private routing backbone
- Fast provisioning (under 30 seconds)
- Transparent, linear pricing
Nirvana gives you the performance of a private data center without the overhead.
Q: What is a hypervisor?
Section titled “Q: What is a hypervisor?”A: A hypervisor is a thin software layer between the physical server and the operating system. It lets you provision, isolate, and manage workloads on the same machine.
Two main types:
- Type 1 (bare metal) — runs directly on the hardware, fast and secure.
- Type 2 (hosted) — runs on top of an OS, adding more overhead.
Hyperscalers use large, multi-tenant Type 1 hypervisors to carve up one physical server into many virtual machines for different customers. That model is efficient for them, but the shared infrastructure introduces scheduling delays, virtualization bloat, and performance jitter — especially under load.
Nirvana Labs also uses Type 1, but single-tenant. Each machine is fully yours — no other workloads compete for resources. The hypervisor is ultra-thin and used only for provisioning, isolation, and automation, not sharing.
Q: What’s special about Nirvana’s hypervisor?
Section titled “Q: What’s special about Nirvana’s hypervisor?”A: Nirvana’s hypervisor is a very thin, single-tenant Type 1 layer designed specifically for Web3 workloads.
While hyperscalers use large multi-tenant hypervisors to share infrastructure — adding overhead, noisy neighbors, and unpredictable performance — Nirvana gives each customer a dedicated machine with near-zero overhead and deterministic control.
What makes it special:
- Single-tenant architecture: No resource contention or noisy neighbors.
- Ultra-lightweight: Near-zero overhead, no multi-tenant bloat.
- Performance-optimized: Tuned for high-CPU blockchain workloads (e.g., sequencers, validators, indexers).
- Hyper-converged design: Compute, storage, and networking unified for predictable low latency.
- Custom drivers and networking: Engineered to minimize jitter and maximize IOPS.
- Deterministic control: You get root access and full machine performance, while the hypervisor quietly handles provisioning, isolation, and automation.
👉 In short: single-tenant, bare-metal performance with cloud-grade flexibility without hyperscaler trade-offs.
Q: Why do you call it a bare metal cloud if there’s a hypervisor?
Section titled “Q: Why do you call it a bare metal cloud if there’s a hypervisor?”A: Most “bare metal” providers today run a lightweight control layer; and so do we.
The key difference is how the hypervisor is used:
Traditional hyperscalers run heavy, multi-tenant hypervisors designed to share infrastructure across many customers.
Nirvana uses a very thin, single-tenant hypervisor that sits directly on the physical server. It isn’t used to share resources between unrelated tenants; it’s there to give you full control of the machine with near-zero overhead.
This design preserves true bare-metal performance while adding the flexibility of clean provisioning, isolation, and automation without compromising the core benefits of dedicated infrastructure.
Q: How are compute, storage, and networking handled?
Section titled “Q: How are compute, storage, and networking handled?”A: Nirvana uses a hyper-converged architecture, meaning compute, storage, and networking are integrated into a single stack rather than split into siloed services. This design eliminates many of the bottlenecks common in traditional cloud setups.
- Low-latency compute: Workloads run on dedicated bare metal servers located near major chain hubs. No multi-tenancy means no noisy neighbors, ensuring deterministic performance even at peak traffic.
- Elastic block storage: Our high-performance NVMe-based storage fabric scales with your workload. It’s optimized for sustained blockchain throughput, giving you fast, reliable data access without shared resource contention.
- Private networking with Nirvana Connect: Instead of routing over the public internet, traffic flows through a private, low-jitter backbone that connects directly to other major clouds and partners. This dramatically reduces latency and egress costs while improving stability.
- End-to-end consistency: Because everything runs on the same node and backbone, throughput stays stable under heavy load — ideal for execution workloads like sequencers, validators, routers, and indexers.
Nirvana removes the layers that slow Web3 workloads down delivering bare-metal performance, elastic scaling, and private network reliability in one stack.
Q: How fast can I deploy infrastructure on Nirvana?
Section titled “Q: How fast can I deploy infrastructure on Nirvana?”A: Typically under 30 seconds. You can deploy clusters near chain hubs, scale quickly, and maintain predictable performance.
Q: How does Nirvana handle monitoring and uptime?
Section titled “Q: How does Nirvana handle monitoring and uptime?”A: Observability is native:
- Real-time performance monitoring
- Built-in alerting & scaling
- Failover and self-healing clusters
Q: What’s the execution flow of RPC calls?
Section titled “Q: What’s the execution flow of RPC calls?”A: It’s the core supply chain of blockchain execution : the flow that powers every wallet, dApp, bridge, and sequencer.
-
Protocols: the source of truth
These are the blockchains themselves (e.g., Ethereum, Arbitrum, Base).
They produce blocks, validate transactions, and maintain the canonical state of the network.
-
Node Providers: the execution backbone
Node providers run blockchain node software (such as Geth, Erigon, or Besu) on high-performance servers, keeping these nodes continuously synced with the live network to maintain the latest state. Beyond syncing, they index and store chain data so applications can query it quickly and reliably. Node providers also ensure uptime and stability by scaling compute, storage, and networking capacity as demand grows, while monitoring and maintaining their infrastructure to handle traffic spikes without downtime. Many colocate their infrastructure near major chain hubs to minimize latency and jitter, enabling faster and more predictable RPC responses for performance-critical workloads.
-
RPC Calls: the interface layer
Every interaction with a blockchain starts with an RPC call; whether fetching data, broadcasting a transaction, or querying state. These calls originate from wallets, dApps, or backend services and are sent to node providers, which act as the interface to the blockchain. The node provider processes the request, queries its synced node for the latest on-chain state, and returns verified data back to the application.
🧠 Execution flow example:
App / Wallet → RPC Call → Node Provider → Protocol → Node Provider → App / Wallet
Q: What’s the difference between Archive Node, Full Node, Light Node, and Dedicated Node?
Section titled “Q: What’s the difference between Archive Node, Full Node, Light Node, and Dedicated Node?”- Archive Node — Stores the entire blockchain history, including all historical states and blocks.
- Full Node — Maintains the current blockchain state and validates all blocks and transactions.
- Light Node — Stores only block headers and relies on full nodes to retrieve detailed data.
- Dedicated Node — A full or archive node allocated exclusively to one customer, with guaranteed performance and resources.
Q: How do you handle data delivery and RPC traffic?
Section titled “Q: How do you handle data delivery and RPC traffic?”A: Nirvana runs its own proprietary CDN and unified API gateway, co-located with compute and storage.
- No Cloudflare/Fastly dependency
- Custom routing and caching
- Dedicated node pools
- Programmable routing for sequencers, validators, and indexers
- Linear pricing with no per-request bloat
Q: Do you support Kubernetes?
Section titled “Q: Do you support Kubernetes?”A: Yes. You can spin up your own Kubernetes cluster (like EKS) on Nirvana today.
Nirvana Kubernetes Service (NKS) — our fully managed Kubernetes offering — is currently in development and will be available soon.
Autoscaler is also part of this upcoming release and remains our #1 feature request.
Q: How does Nirvana compare to hyperscalers?
Section titled “Q: How does Nirvana compare to hyperscalers?”| Category | Hyperscalers | Nirvana |
|---|---|---|
| Infra | Shared + virtualized | Bare metal, private |
| Architecture | Siloed layers | Hyper-converged |
| Provisioning | Slow & complex | Fast, deterministic |
| Observability | Add-on tooling | Built-in |
| Routing | Public internet | Private backbone |
| Pricing | Per GB / per request | Transparent, linear |
| Fit for Web3 | Retrofitted | Purpose-built |