WhiteFiber
WhiteFiber (Nasdaq: WYFI) provides vertically integrated AI infrastructure combining GPU cloud services and Tier III AI-optimized data center colocation, serving AI developers, neoclouds, and regulated enterprises across North America, Europe, and Iceland.
- Company typePublic
- Founded2024
- HeadquartersNew York, United States
- Headcount11–50
- GTM typeB2B
- OfferingServices
What WhiteFiber does
WhiteFiber, Inc. (Nasdaq: WYFI) is a vertically integrated AI infrastructure provider that combines GPU cloud computing with AI-optimized data center colocation. The company operates Tier III facilities at the NC-1 campus in Madison, North Carolina (99 MW Duke Energy agreement, scalable to 200 MW), the MTL-3 facility in Saint-Jérôme, Quebec, and the MTL-1 facility in Montreal, while running cloud services primarily out of a third-party data center in Iceland and securing third-party capacity in France for European expansion. WhiteFiber serves AI developers, ML teams, neoclouds and hyperscalers (notably Nscale, with a 10-year $865M anchor deal), and regulated enterprises in healthcare, finance, and biotech that require high-density, sovereign AI infrastructure.
The platform is engineered around NVIDIA HGX H100, B200, GB200, and upcoming B300/GB300 GPU clusters connected by up to 3.2 Tb/s InfiniBand/RoCE fabric and 800 Gb/s Ethernet, with WEKA/VAST/Ceph storage accelerated by NVIDIA GPUDirect RDMA (up to 40 GB/s single-node). Facilities support up to 150 kW per cabinet with direct-to-chip liquid cooling, 2N power, N+1 cooling, and multi-megawatt scalability to 24 MW+ per site. WhiteFiber is an NVIDIA Preferred Partner and integrates DriveNets Network Cloud-AI Ethernet, Kubernetes/Slurm/Terraform orchestration, and SOC 2 Type II / HIPAA / ISO 27001-aligned compliance posture.
Revenue is generated through three primary streams: cloud GPU services (hourly/monthly billing at $2.50–$3.60/HR for B200/GB200/B300/GB300, plus reserved multi-year contracts), colocation (long-term modified-gross leases with annual escalators — exemplified by the $865M Nscale NC-1 deal), and Private AI managed services. Go-to-market combines enterprise direct sales (separate Cloud Team and Data Center Team), developer self-serve/API access, and marketplace distribution via ShadeForm. FY2025 revenue was $79.2M (+66% YoY) and Q1 2026 was $21.9M (+31% YoY), with the company pre-profitable ($24.7M net loss in 2025) as it invests heavily in capacity buildout against a ~1 GW development pipeline.
WhiteFiber firmographics
Firmographics- Name
- WhiteFiber
- Legal name
- WhiteFiber, Inc.
- Website
- https://whitefiber.com
- Company type
- Public
- Founded year
- 2024
- Operating status
- Operating
- Headcount range
- 11–50 employees
- Short description
- WhiteFiber (Nasdaq: WYFI) provides vertically integrated AI infrastructure combining GPU cloud services and Tier III AI-optimized data center colocation, serving AI developers, neoclouds, and regulated enterprises across North America, Europe, and Iceland.
- Ownership category
- akta.pro rank
WhiteFiber industry classification
Industry- Product category
- AI Infrastructure
- NAICS
- Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services (518), Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services (51821)
- SIC
- Computer Communications Equipment (3576), Services-Computer Processing & Data Preparation (7374), Services-Prepackaged Software (7372)
- akta.pro primary industry
- AI Compute Cloud & GPU-as-a-Service (HDAAAAAK)
- akta.pro secondary industries
- AI Compute Virtualization & Scheduling (GPU virtualization, cluster schedulers) (HDAAAAAG), AI Server Systems & HGX/Accelerator Platforms (HDAAAAAB), AI Networking for Clusters (InfiniBand/Ethernet/RDMA) (HDAAAAAE), AI Interconnects & Fabric (NVLink/CXL/PCIe/Switches) (HDAAAAAD)
Keywords
Where WhiteFiber is headquartered
LocationHeadquarters
- HQ city
- New York
- HQ country
- United States
- HQ region
- North America
Offices7 records
Markets served
WhiteFiber business model
Business model- GTM type
- B2B
- Offering type
- Services
- Cost components
- Infrastructure, Technology or R&D, Supply Chain, Operations, Personnel, Marketing or Sales
Revenue model
- Cloud Services (GPU Cloud): Customers pay for on-demand, reserved, or multi-year access to NVIDIA GPU clusters (HGX H100, B200, GB200, plus B300/GB300) and complementary CPU compute, storage, and networking. Pricing is hourly, monthly, or reserved with 12–24 month commitments. Includes both direct end-user customers and GPU marketplace resellers.
- Colocation Services: Customers lease data center space, power, cooling, and supporting infrastructure inside WhiteFiber's owned Tier III facilities. Includes long-term (multi-year to 10-year) modified-gross colocation leases with annual rate escalators, milestone-based installation payments, PUE/electricity/property-tax passthroughs. Example: 10-year, $865M Nscale deal at NC-1.
- Reserved GPU Capacity (Long-Term Reservations): Multi-year reserved GPU compute instances on Blackwell and Grace Blackwell 200/300 series GPUs with guaranteed bare-metal capacity, custom cluster configuration, and predictable pricing/capacity planning.
- Storage and Networking Add-On Services: Block and shared file system storage via WEKA/VAST/Ceph, billed per GB per month ($0.06/GB/month publicly disclosed), plus high-throughput Ethernet/InfiniBand networking.
- Private AI Managed Services: Turn-key private AI cloud deployments either in WhiteFiber data centers or approved third-party facilities, including cluster architecture design, deployment orchestration, configuration management, day-to-day operations, monitoring, lifecycle upgrades, and compliance/governance alignment.
Pricing tiers
| Model | Billing | Price |
|---|---|---|
| Usage-based | Multi-year contract | NVIDIA B200 — $2.50/HR, 12-month minimum term |
| Usage-based | Multi-year contract | NVIDIA GB200 — $2.55/HR, 12-month minimum term |
| Usage-based | Multi-year contract | NVIDIA B300 — $3.50/HR |
| Usage-based | Multi-year contract | NVIDIA GB300 — $3.60/HR |
| Usage-based | Pay-as-you-go | CPU Compute — Dell PowerEdge R660, Dual Intel 6430 (32 core x 2) |
| Usage-based | Monthly | Storage (WEKA / Vast / Ceph) |
| Subscription | Multi-year contract | Colocation Services (quote-based) |
| Other | Multi-year contract | Private AI / Hybrid AI (quote-based) |
Go-to-market motion4 records
Distribution channels4 records
Marketing channels8 records
WhiteFiber product offering
Product offeringCore offering
WhiteFiber is a vertically integrated AI infrastructure provider that sells AI-optimized data center colocation and GPU cloud services built around NVIDIA HGX H100, B200, GB200, and upcoming B300/GB300 systems. The company owns and operates Tier III facilities in Quebec and North Carolina and also deploys capacity via third-party data centers (Iceland, France), packaging GPU compute, high-throughput storage, InfiniBand/RoCE networking, and turn-key Private AI managed services for enterprise and neocloud customers.
Product overview
WhiteFiber is a vertically integrated AI and HPC infrastructure platform that combines data center colocation and GPU cloud services under a unified engineering stack. At its core, the AI + HPC Colocation and Enterprise Colocation products anchor the company's Tier III facilities (including the flagship NC-1 campus in Madison, North Carolina and the MTL-3 facility in Quebec), while Cloud Compute delivers on-demand NVIDIA GB200, B200, B300, GB300, and HGX H100 GPU capacity complemented by Cloud Storage (WEKA, VAST, Ceph), Cloud Networking (DriveNets-powered), Cloud Observability, and Cloud Security. On top of this infrastructure, the AI Infrastructure, Hybrid AI, and Private AI solutions package the same building blocks into deployable strategies for enterprise customers, with Reserved GPU Cloud Instances providing long-term NVIDIA Blackwell and Grace Blackwell commitments.
Differentiator
Problem solved
Functional benefit
Products and services
- AI + HPC Colocation
Quantifiable outcome
- Retrofit strategy can cut build costs by up to 40% and accelerates time-to-market
- +6 more outcomes
Companies that use WhiteFiber
Customer profileNamed customers5 records
Segments5 records
Ideal customer profiles4 records
WhiteFiber technology and API
TechnologyTechnology focussed Yes
API detail
- Has API
- Yes
- API docs
- API detail
Core technology
AI maturity
App detail
Integration9 records
AI capability7 records
Feature8 records
WhiteFiber partnerships and signals
Strategic signalPartnerships
Twelve partnerships are on record, tiered minor, flagship and core.
- Hyperbolicminor$17M agreement using H200 GPUs from WhiteFiber's existing fleet to support Modal Labs as the end customer, evidencing channel partner relationships for GPU cloud distribution.
- Nscale Global Holdingsflagship10-year master colocation agreement executed via WhiteFiber's subsidiary Enovum Data Centers Corp. for 40 MW of critical IT load at the NC-1 data center campus in Madison, North Carolina, representing approximately $865M in total contract value. Nscale is deploying this capacity to power AI infrastructure for leading global investment-grade technology customers, with potential expansion to double initial size by end of 2027.
- DriveNetscoreWhiteFiber deployment in Iceland was used by DriveNets for live testing demonstrating Network Cloud-AI Ethernet-based solution outperforming Nvidia InfiniBand by 6%, Spectrum-X by 15%, and traditional Ethernet by 72% in NCCL benchmarks. DriveNets is listed as WhiteFiber's networking partner on the partners page.
- Bit Digital CanadacoreWhiteFiber's Q2 2025 financials were reported on a consolidated basis reflecting parent Bit Digital's majority stake; Bit Digital Canada IT Director Jiashu (Bill) Xiong also serves as an Independent Director on WhiteFiber's board, evidencing close cross-entity operational ties in Canada-based data center operations.
- NVIDIAflagshipWhiteFiber is an NVIDIA Preferred Partner, co-branded prominently across the homepage and colocation pages. WhiteFiber's compute platform is built around NVIDIA HGX H100, B200, GB200, and B300/GB300 GPUs, with NVIDIA Quantum InfiniBand fabric and GPUDirect RDMA storage acceleration.
- SupermicrocoreHardware partner listed on the WhiteFiber partners page; Supermicro servers are part of WhiteFiber's GPU cluster deployments.
- Dell TechnologiescoreHardware partner listed on the WhiteFiber partners page. Dell PowerEdge R660 (dual Intel 6430) servers are referenced in the published CPU compute pricing.
- Hewlett Packard Enterprise (HPE)coreHardware partner listed on the WhiteFiber partners page for server and storage infrastructure.
- VAST DatacoreStorage partner providing VAST high-performance storage as one of three tier options on WhiteFiber Cloud ($0.06/GB/month). Co-located in the same data center pods to enable high-throughput distributed storage for AI workloads.
- WEKAcoreStorage partner providing WEKA high-performance file system storage for WhiteFiber Cloud at $0.06/GB/month. Co-located in the same data center pod for high-throughput distributed storage.
- ShadeFormcoreGPU Marketplace partner that distributes WhiteFiber GPU compute to a broad base of developers and ML teams via the ShadeForm marketplace.
- BoosteroidcoreStrategic partnership to finance the global expansion of Boosteroid's cloud gaming infrastructure with $700M in capital dedicated to acquiring custom high-performance servers (engineered by ASUS and powered by AMD technology) across data centers in Europe, North America, and South America. Reduced server acquisition and installation time by 50%.
Scale indicators17 records
Recent moves7 records
Expansion highlights7 records
WhiteFiber competitors and assessment
Company assessmentDirect peers
- CoreWeave: CoreWeave is a leading GPU cloud/neocloud provider offering NVIDIA H100/B200/GB200 capacity to AI developers and enterprises. It is the most direct competitor to WhiteFiber's cloud services, with a similar product mix of bare-metal GPU compute, InfiniBand fabric, and AI-optimized storage.
- Nebius Group: Nebius is a publicly traded AI infrastructure provider offering GPU cloud capacity and full-stack AI cloud services built on NVIDIA hardware. Comparable to WhiteFiber's vertically integrated GPU cloud and colocation platform, particularly for European and international AI customers.
- Crusoe Energy: Crusoe builds vertically integrated AI data centers and GPU cloud services, with a distinctive focus on stranded/renewable power. Like WhiteFiber, it pairs purpose-built data center capacity with NVIDIA GPU clusters and serves both cloud and colocation customers.
- Lambda Labs: Lambda operates GPU cloud instances and private AI clusters on NVIDIA hardware, targeting AI developers and enterprises. Comparable to WhiteFiber's GPU cloud and Private AI offerings, particularly in published hourly GPU pricing and bare-metal access.
Broad incumbents
- Equinix: Equinix is a global colocation and interconnection incumbent with hundreds of IBX data centers. Its enterprise colocation and high-density power offerings overlap with WhiteFiber's enterprise colocation, though Equinix is not AI-specialized and serves a much broader customer base.
- Digital Realty Trust: Digital Realty is a global hyperscale and enterprise colocation provider with extensive Tier III+ data center capacity. Overlaps with WhiteFiber's enterprise colocation offering, though it is broader and less AI-specialized than WhiteFiber's high-density, liquid-cooled footprint.
- Vantage Data Centers: Vantage Data Centers operates large-scale hyperscale and enterprise colocation campuses across North America and Europe. Comparable to WhiteFiber's high-density colocation thesis, with overlap in the Madison/Greensboro and Montreal markets and a similar focus on power-dense GPU tenants.
Emerging players
- Nscale: Nscale is WhiteFiber's largest colocation anchor customer and a neocloud provider building sovereign AI infrastructure. It is a partial peer via its GPU cloud business, which competes with WhiteFiber Cloud in the broader AI infrastructure stack.
- Applied Digital: Applied Digital builds and operates large-scale AI/HPC data centers and offers GPU cloud services, with a similar neocloud/AI-host model. Comparable to WhiteFiber's vertically integrated platform, particularly for high-density GPU hosting at scale.
- Tract (formerly CIM Group / TierPoint-style private developer): Tract is a private developer building AI-optimized data center campuses across the US, similar in scale and intent to WhiteFiber's ~1 GW development pipeline. Comparable as an emerging high-density colocation developer targeting AI/HPC tenants.
Market position
Strengths5 records
Weaknesses5 records
Competitive moat6 records
Key risks6 records
Key highlights7 records
Customer concentration
WhiteFiber social profiles
Digital presenceWhiteFiber compliance and trust
Trust signalCompliance6 records
WhiteFiber financial estimates
Financial estimateRevenue estimate
Valuation estimate
WhiteFiber leadership team
Management profileNumber of profiles
Profiles5 records
WhiteFiber subsidiaries and ownership
Company hierarchySubsidiaries1 record
WhiteFiber funding detail
Funding detailFunding overview
Funding rounds5 records
Investors3 records
Funding detail is available on the Subscription and Enterprise plan.Contact sales →
WhiteFiber M&A and investment
M&A and investmentM&A
Investments
M&A and investment is available on the Subscription and Enterprise plan.Contact sales →
Frequently asked questions about WhiteFiber
What does WhiteFiber do?
WhiteFiber is a vertically integrated AI infrastructure provider that sells AI-optimized data center colocation and GPU cloud services built around NVIDIA HGX H100, B200, GB200, and upcoming B300/GB300 systems. The company owns and operates Tier III facilities in Quebec and North Carolina and also deploys capacity via third-party data centers (Iceland, France), packaging GPU compute, high-throughput storage, InfiniBand/RoCE networking, and turn-key Private AI managed services for enterprise and neocloud customers.
Is WhiteFiber a public or private company?
WhiteFiber is a public company. It is classified as public and is currently operating.
When was WhiteFiber founded?
WhiteFiber was founded in 2024. It employs 11 to 50 people.
Where is WhiteFiber based?
WhiteFiber is headquartered in New York, United States, in the North America region.
How does WhiteFiber make money?
Five revenue lines are on record. Cloud Services (GPU Cloud) is the primary driver. The others are colocation Services, reserved GPU Capacity (Long-Term Reservations), storage and Networking Add-On Services and private AI Managed Services.
Who are WhiteFiber's main competitors?
Direct peers on record are CoreWeave, Nebius Group, Crusoe Energy and Lambda Labs. Broad incumbents are Equinix, Digital Realty Trust and Vantage Data Centers. Emerging players are Nscale, Applied Digital and Tract (formerly CIM Group / TierPoint-style private developer).
Does WhiteFiber have an API?
Yes. WhiteFiber offers developer-friendly API/CLI tooling to deploy and manage AI workloads across its data centers. The hybrid platform explicitly supports integration with REST APIs, Terraform for infrastructure-as-code, and CI/CD workflows. Customers can orchestrate workloads via Kubernetes, Slurm, or custom stacks and manage both bare metal and virtualized instances via the API tooling.
What industry is WhiteFiber in?
WhiteFiber's product category is AI Infrastructure. Its primary akta.pro industry code is HDAAAAAK, AI Compute Cloud & GPU-as-a-Service, with a secondary code of HDAAAAAG, AI Compute Virtualization & Scheduling (GPU virtualization, cluster schedulers). Its NAICS code is 518 and its SIC code is 3576.