What Verda (formerly DataCrunch) does

Verda (legally DataCrunch Oy) is a Helsinki-based, venture-backed European AI cloud hyperscaler operating data centers in the Nordics on 100% renewable energy. The company operates a vertically integrated full-stack platform spanning data-center foundation, core infrastructure (NVIDIA GPUs with NVLink, InfiniBand, and RoCE; CPU nodes; block, shared filesystem, and object storage), cloud compute (GPU instances, Instant Clusters of 16x–128x GPUs, bare-metal clusters), an inference platform (auto-scaling Serverless Containers, Batch Jobs, Container Registry, Inference API), and an AI developer ecosystem (web console, REST API, Python SDK, MCP server, Terraform/OpenTofu providers, plus dstack and SkyPilot integrations). It holds NVIDIA Preferred Partner status with one of Europe's earliest deployments of NVIDIA GB300 NVL72 and HGX B300, and an in-house AI Lab conducts research on large-scale training, RL frameworks, inference optimization, and compilers/kernels.

The product set includes GPU Instances, Instant Clusters, Bare-metal Clusters, Serverless Containers, Managed Endpoints (hosted FLUX.1 Kontext dev/pro/max, FLUX.1 Krea, FLUX.1 [dev], FLUX.2, FLUX.2 [klein], and Whisper audio), and a Confidential Computing add-on combining AMD SEV-SNP CPU memory encryption with NVIDIA Confidential Computing on RTX PRO 6000 Blackwell, with hardware-attested TEE and encrypted PCIe transfers. The revenue model is primarily usage-based (on-demand GPU compute, dynamic pricing, long-term discounts, pay-per-output inference at $0.025–$0.08/image for FLUX endpoints), with enterprise contracts for bare-metal, co-development, and confidential computing PoCs. Customers span frontier model developers (Black Forest Labs, 1X), AI startups (Freepik, Unbabel, Magnific, Simli, Happywhale), large enterprises (Sony, Nokia, NEC, ExpressVPN), and universities (Harvard, MIT, Korea University).

Verda is cash-flow positive as of Q1 2026 with annualized revenue exceeding $60M (€51.3M), and is positioning itself as a sovereign, GDPR-native, lower-cost European alternative to AWS, Azure, and Google Cloud for AI workloads, including through a partnership with Supermicro (May 2026) and Compal (May 2026) for rack-scale Blackwell deployment across Europe, the US, and Asia, and an EU AI GigaFactory proposal for Latvia.

Name

Verda (formerly DataCrunch)

Legal name

DataCrunch Oy

Website

https://verda.com

Company type

Private

Founded year

2021

Operating status

Operating

Headcount range

251–500 employees

Short description

Verda (formerly DataCrunch) is a Helsinki-based European AI cloud hyperscaler offering vertically integrated GPU compute, instant clusters, serverless inference, and confidential computing on NVIDIA Blackwell hardware. It serves frontier model developers, AI startups, regulated enterprises, and research institutions as a GDPR-native alternative to US hyperscalers.

Ownership category

akta.pro rank

Product category

AI Cloud Infrastructure

NAICS

Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services (5182), Computer Facilities Management Services (541513)

SIC

Services-Computer Programming, Data Processing, Etc. (7370), Services-Computer Processing & Data Preparation (7374)

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), Private Cloud Networking & SDN/NFV Management (HDABABAK), Data Center Operations & Managed Services (Remote Hands, NOC/SOC) (HDABANAA)

Keywords

• AI cloud infrastructure
• GPU cloud computing
• Serverless inference platform
• Confidential computing services
• Managed inference endpoints

Headquarters

HQ city

Helsinki

HQ country

Finland

HQ region

Europe

Offices4 records

Markets served

GTM type

B2B

Offering type

Software

Cost components

Technology or R&D, Infrastructure, Personnel, Operations, Marketing or Sales

Revenue model

1. Usage-based GPU compute: On-demand GPU instances, Instant Clusters, and Bare-metal Clusters billed based on consumption with dynamic pricing, plus pay-as-you-go serverless containers and managed inference endpoints billed per request or output.
2. Managed inference API per-output: Managed Endpoints charge per output unit (e.g., FLUX.1 Kontext dev at $0.025/image, pro at $0.04/image, max at $0.08/image) with enterprise discounts available.
3. Long-term/contracted enterprise agreements: Enterprise and co-development engagements with long-term contracts; customers can enter multi-year terms that require contacting support to terminate.
4. Co-development professional services: Custom full-stack AI solutions built and maintained by Verda's engineers, including confidential computing engagements, for customers requiring bespoke infrastructure.
5. Self-serve credit-based billing: Customers can purchase compute credits or pay by credit/debit card or invoice via Stripe, with optional recurring, top-ups, renewal, or usage-based charges.

Pricing tiers

Go-to-market motion4 records

Distribution channels5 records

Marketing channels8 records

Core offering

Verda (legally DataCrunch Oy) operates a vertically integrated European AI cloud platform built on NVIDIA GPUs, providing on-demand GPU Instances, Instant Clusters (16x-128x GPUs with InfiniBand), Bare-metal Clusters, Serverless Containers, Managed Endpoints for inference, and Confidential Computing with hardware-attested Trusted Execution Environments. Customers access the platform via a self-serve web console, REST API, Python SDK, and Terraform/OpenTofu providers, with services sold on usage-based pay-as-you-go, per-output, or long-term contract terms from Nordic data centers powered by 100% renewable energy.

Product overview

Verda (formerly DataCrunch) is a vertically integrated European AI cloud platform — not a single product — built as a full-stack architecture spanning data center foundation, core infrastructure (NVIDIA GPUs, CPU nodes, block storage, shared filesystem, object storage, NVLink, InfiniBand, RoCE), cloud compute (Instant Clusters, virtual machines), an inference platform (auto-scaling containers, batch jobs, container registry, Inference API), an AI developer ecosystem (web console, API, SDK, Terraform, partner integrations), and an in-house AI Lab. Customers access compute through GPU Instances (1x-8x GPUs with NVLink) for entry-level workloads, scale up via Instant Clusters (16x-128x GPUs with InfiniBand) or Bare-metal Clusters for custom deployments, and serve models through Serverless Containers and Managed Endpoints (including hosted FLUX.1 Kontext dev/pro/max, FLUX.1 Krea, FLUX.1 [dev], FLUX.2, FLUX.2 [klein], and Whisper). Confidential Computing is an emerging add-on layer that wraps selected NVIDIA Blackwell GPUs in hardware-attested Trusted Execution Environments. All of this is operated from Nordic data centers running on 100% renewable energy and is positioned as a sovereign, GDPR-compliant alternative to US hyperscalers, with Verda holding NVIDIA Preferred Partner status.

Differentiator

Problem solved

Functional benefit

Products and services

• GPU Instances
• Instant Clusters
• Bare-metal Clusters
• Serverless Containers
• Managed Endpoints
• Confidential Computing
• Co-development
• FLUX.1 Kontext (Managed Endpoint)
• Verda API & Python SDK

Quantifiable outcome

• Up to 90% lower GPU compute cost vs hyperscalers
• +5 more outcomes

Named customers23 records

Segments5 records

Ideal customer profiles4 records

Technology focussed Yes

API detail

Has API

Yes

API docs

API detail

Core technology

AI maturity

App detail

Integration5 records

AI capability13 records

Feature9 records

Partnerships

Eleven partnerships are on record, tiered minor, flagship and core.

• Wave VenturesminorGTM or Marketing Partner · 17 June 2026Verda is among the supporters (alongside AWS, Anthropic, Lovable, Modal) of Wave Ventures' Wave Rebellion Grant providing €240,000 in funding and partner credits to 10 young Nordic builders.
• SupermicroflagshipTechnology or Integration · 27 May 2026Verda selected Supermicro's NVIDIA Blackwell-accelerated, rack-scale, liquid-cooled systems (GB300 NVL72, HGX B300/B200, RTX PRO 6000) to power full-stack AI cloud infrastructure across Europe, the US, and Asia, enabling self-service AI instances, serverless containers, and managed inference endpoints.
• Compal ElectronicscoreStrategic or Co-development Partner · 7 May 2026Strategic partnership under which Compal supplies high-density, liquid-cooled AI server platforms to accelerate Verda's AI infrastructure build-out across Europe and the APAC region, designed for agentic AI workloads with high concurrency and thermal efficiency.
• ExpressVPNflagshipStrategic or Co-development Partner · 16 April 2026Joint R&D with ExpressVPN to enable confidential computing on NVIDIA Blackwell architecture for the ExpressAI privacy-first AI platform; enabled confidential computing for multi-node configurations on NVIDIA HGX B200 and worked on early-stage distributed secure GPU clusters.
• NVIDIAflagshipTechnology or Integration · 1 January 2026Verda is an NVIDIA Preferred Partner within the NVIDIA Partner Network (NPN), with some of the earliest deployments of NVIDIA Blackwell Ultra platforms in Europe including GB300 NVL72 and HGX B300. Joint engineering on confidential computing for NVIDIA Blackwell architecture.
• Latvian alliance (incl. Latvian government)flagshipStrategic or Co-development Partner · 1 August 2025DataCrunch-Latvian alliance submitted a proposal for the EU's AI GigaFactory initiative to develop an AI gigafactory in Latvia providing renewable-powered AI compute capacity.
• Black Forest LabscoreTechnology or IntegrationBlack Forest Labs' FLUX.1 Kontext (dev, pro, max) and FLUX.1 Krea, FLUX.2 image generation models are hosted as Managed Endpoints on Verda infrastructure, with FLUX.1 Kontext max run on Black Forest Labs' infrastructure and accessed via Verda.
• WaveSpeedAIcoreTechnology or IntegrationWaveSpeedAI accelerates FLUX.1 Kontext inference on Verda infrastructure, enabling faster, cheaper inference with no loss in quality for managed endpoints.
• vLLMcoreTechnology or IntegrationvLLM listed as customer/partner and integrated into Verda's serverless containers and Instant Clusters; Verda publishes a tutorial for deploying vLLM inference on Instant Cluster using Ray.
• SGLangcoreTechnology or IntegrationVerda sponsors SGLang with compute resources and infrastructure support; SGLang's research on RL training with FP8 and INT4 utilizes NVIDIA Hopper, Blackwell, and Blackwell Ultra platforms on Verda.
• 1XcoreStrategic or Co-development PartnerVerda collaborates with 1X on building multi-GPU inference for the 1XWM generative video model; an in-house Verda team (Revontuli) won both tracks of the 1X World Model Challenge using Verda Instant Clusters with Blackwell GPUs and InfiniBand.

Scale indicators15 records

Recent moves8 records

Expansion highlights7 records

Direct peers

• Fluidstack: GPU cloud provider operating large-scale NVIDIA fleets (including Blackwell) for AI training and inference customers. Direct peer in large-cluster GPU compute and enterprise AI customer base.
• Together AI: AI cloud platform combining GPU clusters with managed inference APIs for open-source models. Direct peer in the inference-platform and open-source-AI developer ecosystem that overlaps with Verda's SGLang/vLLM partnerships.
• CoreWeave: US-based GPU cloud provider offering NVIDIA-powered instances, clusters, and managed inference. Direct peer in vertically integrated AI cloud with similar product breadth (instances, clusters, serverless, managed inference) and target customer segments.
• Lambda Labs: US-based GPU cloud provider offering on-demand NVIDIA H100/B200 instances, clusters, and dedicated hardware for AI training and inference. Direct peer in developer-first GPU-as-a-service.
• Crusoe Energy: US-based GPU cloud provider pairing AI compute infrastructure with sustainable/stranded-energy data centers. Direct peer in NVIDIA GPU cloud with a sustainability-led value proposition similar to Verda's renewable-energy and heat-reuse narrative.
• Nebius: AI infrastructure company with European origins operating GPU clusters and cloud services globally. Direct peer in vertically integrated AI cloud positioning, with a comparable mix of training and inference workloads.
• Scaleway: European cloud provider offering GPU instances, clusters, and managed services with a strong French/European sovereignty story. Direct peer in sovereign European cloud competing for the same regulated EU workloads.
• OVHcloud: European cloud provider with GPU instances, bare-metal servers, and managed services targeted at European sovereignty-conscious customers. Direct peer in sovereign EU cloud and infrastructure-as-a-service.

Emerging players

• RunPod: GPU cloud provider offering serverless endpoints, instant clusters, and on-demand instances targeting AI developers. Emerging peer with overlapping developer-first product motion and inference platform features.

Broad incumbents

• Paperspace (DigitalOcean): GPU cloud service (now part of DigitalOcean) offering on-demand NVIDIA instances and notebooks for AI developers. Broad incumbent peer offering overlapping GPU compute as part of a wider cloud portfolio.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat5 records

Key risks5 records

Key highlights7 records

Customer concentration

• Xx.com/verdacloud
• LinkedInlinkedin.com/company/home

Compliance7 records

Revenue estimate

Valuation estimate

Number of profiles

Profiles4 records

Funding overview

Funding rounds17 records

Investors17 records

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Verda (formerly DataCrunch)