Acasia
Acasia is a private US-based AI infrastructure company providing dedicated GPU cloud, managed inference, and a compute marketplace serving AI companies, developers, and investor-family-office clients across 12 US metro data center locations.
- Company typePrivate
- Founded-
- HeadquartersThe Woodlands, United States
- Headcount11–50
- GTM typeB2B
- OfferingSoftware
What Acasia does
Acasia (formerly Acacia) is a private US-based AI infrastructure company that provides dedicated GPU cloud infrastructure and managed inference services for AI workloads. The company operates a vertically integrated platform spanning bare metal GPU compute, orchestration (Fabric GMS), managed inference (ARIA), a compute marketplace (AIX), and the AOS unified operating system. Its hardware footprint includes NVIDIA-certified clusters across 12 US metropolitan data center locations, with notable capacity including an 18 MW Dallas facility housing a 2,048-GPU B300 cluster, a modular Houston data center capable of supporting up to 12,000 Blackwell GPUs, and a Santa Clara colocation site. Core partners include NVIDIA, VAST Data, WEKA, Dell, Lenovo, and Supermicro.
The company serves three primary customer segments: AI companies and large enterprises requiring dedicated single-tenant GPU clusters, developers and AI builders needing self-serve compute access, and investors and family offices participating in a GPU-as-a-Service program that targets 25-35% IRR returns. Revenue mechanics include usage-based hourly GPU rentals (ranging from $1.25/hr for A100 to $5.90/hr for GB300 NVL72), multi-year reserved cluster offtake agreements, managed data center colocation, and managed services for investor-owned GPU fleets. Stated vertical priorities include life sciences and financial services, with demonstrated AI applications in healthcare (Patient Advocate Agents).
The company is led by CEO Richard Cant, with a team assembled through multiple 2026 executive appointments including CTO Alexander De Ridder, CRO Gary Haymann, COO Christopher Lawrence, CCO Matt Renner, Head of AI Brian Rogers, and VP Compute Sales Jeff Kafka. The firm is headquartered in The Woodlands, Texas (with additional executive presence in Houston and Minneapolis), operates with 11-50 employees, and rebranded from Acacia to Acasia in April 2026 as part of a strategic repositioning around accessibility of AI infrastructure.
Acasia firmographics
Firmographics- Name
- Acasia
- Legal name
- Acasia Technology, Inc.
- Website
- https://acasia.com
- Company type
- Private
- Operating status
- Operating
- Headcount range
- 11–50 employees
- Short description
- Acasia is a private US-based AI infrastructure company providing dedicated GPU cloud, managed inference, and a compute marketplace serving AI companies, developers, and investor-family-office clients across 12 US metro data center locations.
- Ownership category
- akta.pro rank
Acasia industry classification
Industry- Product category
- AI Infrastructure / GPU Cloud Computing
- NAICS
- Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services (518), Computer Systems Design and Related Services (5415)
- SIC
- Services-Computer Rental & Leasing (7377), Electronic Computers (3571), Services-Computer Programming, Data Processing, Etc. (7370)
- akta.pro primary industry
- AI Compute Cloud & GPU-as-a-Service (HDAAAAAK)
Keywords
Where Acasia is headquartered
LocationHeadquarters
- HQ city
- The Woodlands
- HQ country
- United States
- HQ region
- North America
Offices6 records
Markets served
Acasia business model
Business model- GTM type
- B2B
- Offering type
- Software
- Cost components
- Technology or R&D, Infrastructure, Personnel, Supply Chain, Marketing or Sales, Operations
Revenue model
- GPU Compute Rentals (Marketplace): On-demand and spot GPU rental through the Acasia Marketplace, where customers rent compute by the hour for short-term or burst needs. Pricing is transparent and per-GPU/hour, with verified bare metal availability across multiple metro regions.
- Reserved GPU Clusters (Dedicated Bare Metal): Dedicated single-tenant GPU clusters — NVIDIA GB300 NVL72, B300, and Vera Rubin — deployed in 3-4 months with guaranteed RFS dates, contracted under multi-year offtake, lease-to-own, or straight rent structures. Power and hardware bundled under one accountable contract.
- GPU-as-a-Service for Investors: Full-service program helping investors purchase GPU fleets: procurement and financing, deployment and managed operations, guaranteed offtake, and marketplace distribution through Acasia Cloud. Typically targeting 25-35% IRR for qualified family offices and funds.
- Data Center Capacity (Colocation and Turnkey): Modular and colo data center capacity sold as contracted power and space (e.g., 9.5 MW tranches, 625 kW in Santa Clara), with optional turnkey cluster procurement and managed operations for customers bringing their own hardware.
Pricing tiers
| Model | Billing | Price |
|---|---|---|
| Usage-based | Pay-as-you-go | A100 — Ampere generation, best price per hour for fine-tuning |
| Usage-based | Pay-as-you-go | H100 — Hopper generation, training workhorse on demand |
| Usage-based | Pay-as-you-go | H200 — Highest memory bandwidth Hopper, fully rented |
| Usage-based | Pay-as-you-go | B300 — Blackwell Ultra HGX, air-cooled |
| Usage-based | Pay-as-you-go | GB300 NVL72 — Grace Blackwell Ultra, liquid-cooled rack-scale |
| Freemium | Pay-as-you-go | AIX Marketplace New User Credit |
| Outcome Based/ Performance | Multi-year contract | GPU-as-a-Service for Investors — IRR Target |
| Subscription | Multi-year contract | Reserved Cluster Offtake — multi-year committed |
Go-to-market motion3 records
Distribution channels6 records
Marketing channels8 records
Acasia product offering
Product offeringCore offering
Acasia provides dedicated, single-tenant GPU cloud infrastructure (NVIDIA GB300 NVL72, B300, H200, H100, A100, Vera Rubin), a managed inference layer (ARIA), a GPU orchestration control plane (Fabric GMS / AOS), and a global GPU compute marketplace (AIX) that aggregates idle and reserved capacity for on-demand and reserved rental. Customers range from AI companies and large enterprises buying reserved clusters to developers and investors accessing on-demand GPU compute or financing GPU fleets with managed operations.
Product overview
Acasia is a vertically integrated AI infrastructure company offering a unified platform architecture with multiple interconnected products. The core platform consists of Acasia Cloud Platform providing unified GPU compute management, paired with the Acasia GPU Marketplace for on-demand and reserved GPU rentals. For dedicated capacity, they offer Reserved GPU Clusters with NVIDIA platforms (Vera Rubin, GB300 NVL72, B300). Supporting products include AIX GMS for GPU resource orchestration and payments, AIX Marketplace for aggregating idle/reserved capacity (up to 90% cost savings), and ARIA for managed inference with intelligent routing. The AOS Platform provides the managed inference control plane with deployment planning and operational controls. Additional offerings include Acasia NEO Cloud for enterprise GPU services and GPU as a Service for investor fleets. The platform serves developers and enterprises across Life Sciences, Financial Services, Travel & Hospitality, Manufacturing, Retail & E-Commerce, and Research & Academia.
Differentiator
Problem solved
Functional benefit
Brands
- AIX Marketplace: Global GPU compute marketplace aggregating idle and reserved capacity with up to 90% cost savings versus traditional cloud providers.
- Acasia NEO Cloud
- Fabric GMS
- AOS Platform
- ARIA
Products and services
- Acasia Cloud Platform
- Acasia GPU Marketplace
- Reserved GPU Clusters Dedicated, single-tenant GPU clusters sourced, financed, powered, and operated end-to-end for AI cloud providers and enterprises; includes reserved allocation of NVIDIA Vera Rubin, GB300 NVL72, and B300 platforms deployed in 3-4 months with guaranteed RFS dates.
- Acasia NEO Cloud Enterprise GPU data-center and infrastructure services powered by NVIDIA for organizations requiring dedicated, auditable infrastructure with NVIDIA-certified reference architectures.
- AIX GMS (GPU Management System) GPU management system for orchestrating compute resources, payments, and pricing intelligence across the marketplace and cluster fleet.
- AIX Marketplace Global GPU compute marketplace aggregating idle and reserved capacity, delivering up to 90% cost savings versus traditional cloud providers with verified bare-metal GPUs, guaranteed power, and turn-key clusters.
- ARIA (AI Runtime Intelligence Architecture) Managed inference layer providing intelligent model routing, outcome-centric execution, real-time reliability, and continuous learning and optimization for production AI and agentic workloads.
- Fabric GMS Control plane for managing GPU infrastructure at scale, providing provisioning, governance, coordination, and packaging aligned with enterprise requirements and economic realities across distributed clusters.
- AOS Platform Unified AI operating system integrating GPU compute, data-center infrastructure, GPU management, and developer access; includes an inference endpoint deployment planner and live health/spend governance dashboard with p50/p95 latency targeting and budget-aware controls.
- GPU-as-a-Service for Investors Full-service investment program for GPU fleets targeting 25-35% IRR with procurement and financing, deployment and managed operations, guaranteed offtake, and marketplace distribution through Acasia Cloud.
Quantifiable outcome
- Up to 90% cost savings vs traditional cloud providers through AIX Marketplace idle capacity aggregation
- +4 more outcomes
Companies that use Acasia
Customer profileSegments7 records
Ideal customer profiles5 records
Acasia technology and API
TechnologyTechnology focussed Yes
API detail
- Has API
- Yes
- API docs
- API detail
Core technology
AI maturity
App detail
AI capability8 records
Feature7 records
Acasia partnerships and signals
Strategic signalPartnerships
Six partnerships are on record, tiered core and major.
- NVIDIAcoreNVIDIA is the primary technology partner providing all GPU hardware — GB300 NVL72, B300, H200, H100, A100, Vera Rubin. All Acasia clusters are NVIDIA certified to reference architecture. NVIDIA Certified Professional AI Networking designation maintained. Partnership encompasses GPU procurement, reference architecture validation, certification, and joint go-to-market for NVIDIA-powered GPU infrastructure.
- VAST DatacoreVAST Data is the preferred parallel filesystem partner for Acasia cluster deployments, tiered for checkpoint throughput, active dataset performance, and long-term retention. Available across GB300 NVL72, B300, and Vera Rubin cluster configurations.
- WEKAcoreWEKA is an alternative parallel filesystem option alongside VAST Data for Acasia cluster deployments, providing tiered storage for checkpointing and active datasets. Offered as a configurable storage option during cluster offtake.
- LenovomajorLenovo is one of three approved OEM hardware partners for Acasia clusters (alongside Dell and Supermicro). Customers can select Lenovo servers for cluster configurations, allowing platform, support contract, and lead time to be matched to customer standards.
- DellmajorDell is one of three approved OEM hardware partners for Acasia GPU clusters (alongside Lenovo and Supermicro). Dell partnership includes hardware procurement and support contracts, with Dell mentioned alongside Acasia's support model combining in-house capabilities with Dell's broader engineering resources.
- SupermicromajorSupermicro is one of three approved OEM hardware partners for Acasia clusters (alongside Dell and Lenovo). Customers can specify Supermicro servers as part of their cluster BOM customization during the offtake process.
Scale indicators13 records
Recent moves7 records
Expansion highlights6 records
Acasia competitors and assessment
Company assessmentMarket position
Competitive moat6 records
Key risks5 records
Key highlights6 records
Customer concentration
Acasia social profiles
Digital presenceAcasia compliance and trust
Trust signalCompliance1 record
Acasia financial estimates
Financial estimateRevenue estimate
Valuation estimate
Acasia leadership team
Management profileNumber of profiles
Profiles7 records
Acasia funding detail
Funding detailFunding overview
Funding rounds
Investors
Funding detail is available on the Subscription and Enterprise plan.Contact sales →
Acasia 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 Acasia
What does Acasia do?
Acasia provides dedicated, single-tenant GPU cloud infrastructure (NVIDIA GB300 NVL72, B300, H200, H100, A100, Vera Rubin), a managed inference layer (ARIA), a GPU orchestration control plane (Fabric GMS / AOS), and a global GPU compute marketplace (AIX) that aggregates idle and reserved capacity for on-demand and reserved rental. Customers range from AI companies and large enterprises buying reserved clusters to developers and investors accessing on-demand GPU compute or financing GPU fleets with managed operations.
Is Acasia a public or private company?
Acasia is a private company. It is classified as founder individual operated bootstrapped and is currently operating.
When was Acasia founded?
Acasia was founded in -1. It employs 11 to 50 people.
Where is Acasia based?
Acasia is headquartered in The Woodlands, United States, in the North America region.
How does Acasia make money?
Four revenue lines are on record. GPU Compute Rentals (Marketplace) is the primary driver. The others are reserved GPU Clusters (Dedicated Bare Metal), GPU-as-a-Service for Investors and data Center Capacity (Colocation and Turnkey).
Does Acasia have an API?
Yes. Acasia provides a developer portal with API keys and Terraform-ready endpoints for provisioning GPU instances, pulling telemetry, managing keys, and reconciling spend. The platform supports CLI and API access for launching instances and managing clusters. No additional API documentation URL or SDK details specified in available sources. Developer documentation is at docs.acasia.com/docs.
What industry is Acasia in?
Acasia's product category is AI Infrastructure / GPU Cloud Computing. Its primary akta.pro industry code is HDAAAAAK, AI Compute Cloud & GPU-as-a-Service. Its NAICS code is 518 and its SIC code is 7377.