Super X AI

Name

Super X AI

Legal name

SuperX AI Technology Limited

Website

https://superx.sg

Company type

Public

Founded year

2025

Operating status

Operating

Headcount range

11–50 employees

Ownership category

akta.pro rank

Product category

AI Infrastructure Hardware

NAICS

Computer and Peripheral Equipment Manufacturing (33411), Computer Systems Design and Related Services (54151), Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services (5182)

SIC

Electronic Computers (3571), Services-Computer Integrated Systems Design (7373)

akta.pro primary industry

GPU-Accelerated & AI Training/Inference Servers (HDACABAG)

Keywords

• AI infrastructure solutions
• AI server hardware
• Data center power systems
• Liquid cooling solutions
• Modular data centers

Headquarters

HQ city

Singapore

HQ country

Singapore

HQ region

Asia

Offices4 records

Markets served

GTM type

B2B

Offering type

Hardware or Manufacturing

Cost components

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

Revenue model

1. AI Hardware Sales: Sale of high-performance AI servers (B200, B300, GB300 NVL72 systems), AI rack systems, and enterprise workstations. Revenue recognized upon delivery of hardware. The company reported $3.6M in trailing twelve-month revenue with 10.16% gross profit margin, with $910M in secured purchase orders and $2.1B in non-binding MOUs indicating significant pipeline conversion expected.
2. AI Infrastructure Solutions (Full-Stack): Integrated solutions combining servers, power systems (800VDC), liquid cooling, and modular factory deployments. Revenue includes system integration, solution design, and deployment services for AI data centers.
3. AI Cloud Services: AI inference cloud services offered through the Denver AI Inference Cloud Hub and AI Cloud platform. Provides on-demand GPU access and reserved capacity models for AI workloads. Emerging revenue stream as company expands cloud capabilities globally.
4. Power & Cooling Infrastructure: Sale of 800VDC power solutions (Panama-800VDC, Aurora-800VDC), liquid cooling systems (CDU, Cryopod), and related power infrastructure products for AI data centers.
5. Technical Services & Maintenance: End-to-end operations and maintenance services for AI infrastructure, supported by 24/7 global support center in Singapore. Reinforces customer retention and provides recurring service revenue.

Pricing tiers

Go-to-market motion2 records

Distribution channels4 records

Marketing channels5 records

Core offering

SuperX AI is a full-stack AI infrastructure solutions provider that designs, manufactures, and delivers high-performance AI servers, 800VDC power systems, liquid cooling solutions, modular data centers, and AI cloud services. The company integrates NVIDIA GPU architectures (HGX B200/B300, GB300 NVL72) with proprietary power, cooling, and modular factory systems for hyperscale AI operators, sovereign AI projects, research institutions, and enterprises. Products are produced at a Global Supply Center in Japan with 20,000 server annual capacity and deployed globally through direct sales and joint ventures.

Product overview

SuperX AI is a full-stack AI infrastructure solutions provider offering a comprehensive portfolio of proprietary hardware, software, and services for AI data centers. The product portfolio includes: AI Servers (XI6150, B300, B200, B200 Standard, NVIDIA GB300 GB300 NVL72 System), AI Workstations (XK1140 Workstation Standard, XK1270 Workstation Ultra), Digital Power Solutions (Panama-800VDC, Aurora-800VDC, GreenPort, Energy Vault, Powercore), Liquid Cooling (Cryopod, Micro-Modular Data Center), Cloud Services (AI Cloud Platform, AI Innovation Centre, SuperX Chat), and the Modular AI Factory. These products work together as an integrated full-stack solution spanning from hardware (GPU servers, power systems, cooling) through software (AI cloud, multi-model platforms) to services (AI Innovation Centre, global supply center), enabling enterprises to deploy AI infrastructure from pilot to production scale.

Differentiator

Problem solved

Functional benefit

Brands

• SuperX APEX AI: Joint venture focused on AI solutions
• SuperX Digital Power
• SuperX Global Service
• SuperX Microinference
• SuperX Optical Communications
• SuperX Cooltech

Products and services

• XI6150 A high-density multi-engine compute platform designed for next-generation AI workloads, supporting flexible GPU topologies with up to 600W per GPU for large-scale inference and training.
• B300 AI Server An 8U dual-processor AI server powered by NVIDIA HGX B300 platform with 8 NVIDIA Blackwell GPUs, featuring 1.8 TB/s GPU-to-GPU bandwidth via 5th-Gen NVLink and NVSwitch for large-scale AI training, inference, and HPC workloads.
• B200 AI Server A 10U AI inference server powered by NVIDIA HGX B200 with 8 NVIDIA Blackwell B200 GPUs connected via 5th-Gen NVLink, 1,440 GB HBM3E memory, and 6th-Gen Intel Xeon processors for peak AI performance.
• B200 Standard Multi-Model Server Enterprise-grade multi-model server with 8 NVIDIA Blackwell B200 GPUs with 192GB HBM3e VRAM, supporting up to 180 concurrent users for enterprise content production, customer service Q&A, and knowledge base applications.
• NVIDIA GB300 NVL72 System Rack-scale AI supercomputing platform delivering up to 1.8 exaFLOPS of FP4 AI compute in a single liquid-cooled rack using NVIDIA GB300 Grace Blackwell Ultra Superchip with 72 Blackwell Ultra GPUs and 36 Grace CPUs, incorporating 800VDC power infrastructure and 800Gb/s InfiniBand connectivity.
• XK1140 Workstation Standard Entry-level AI workstation with 1-2 NVIDIA GeForce RTX 5090 GPUs (32GB GDDR7 VRAM), Intel Core i9-14900KF processor, designed for developer AI proof-of-concept and SME lightweight knowledge base applications.
• XK1270 Workstation Ultra High-performance AI workstation with 4 NVIDIA GeForce RTX 5090 GPUs (32GB GDDR7 VRAM), AMD Ryzen Threadripper processor, supporting up to 80 concurrent users for real-time collaborative scenarios.
• Panama-800VDC A 10kV AC-to-DC power system that streamlines data center power distribution through innovative electrical and magnetic circuit integration, featuring up to 98.5% conversion efficiency, modular design, and deployment capability from 2.5MW to 10MW for new-build AI data centers.
• Aurora-800VDC An 800V DC power system with 60kW high-power modules designed for next-generation 800V AI servers, achieving 98.5% system efficiency with constant 800VDC output, supporting up to 1.2MW per 800mm wide cabinet for retrofitting existing facilities.
• GreenPort A fully integrated power shelter combining 10/6.6 kV medium-voltage switchgear, DC operation power cabinets, integrated monitoring, power distribution, and control systems with complete environmental and safety features for modern data center deployment.
• Energy Vault An integrated HVDC energy storage system eliminating need for UPS, combining batteries, switch boxes, monitoring, power distribution, and environmental controls with independent space and fire protection for safe, reliable stand-alone operation.
• Powercore Diesel power cabinets with control systems and power distribution, featuring standard 40-foot container main compartment and comprehensive auxiliary equipment including parallel control, hydraulic system control, fire protection, and environmental controls.
• Cryopod A dry air cooler for modern data centers providing efficient and reliable heat rejection with year-round natural cooling capability, ultra-low energy consumption, and support for 24/7 continuous operation with service life of 10+ years, available in models from 680kW to 900kW.
• Micro-Modular Data Center (MDC) A compact, prefabricated data center solution integrating IT racks, power distribution, cooling, fire protection, and intelligent monitoring into a single unit for rapid deployment with PUE as low as 1.2.
• Modular AI Factory A modular data center solution integrating compute, cooling, and power systems designed to deploy AI infrastructure in under six months. Offers up to 20MW per module supporting up to 144 NVIDIA GB200 NVL72 systems, achieves 98.5% end-to-end power efficiency and PUE as low as 1.15.
• AI Cloud Platform A cloud platform providing instant access to high-performance NVIDIA GPUs for accelerated AI development lifecycle with flexible, developer-friendly services, including AI inference capabilities and on-demand GPU resource access.
• AI Innovation Centre A joint initiative with ST Telemedia Global Data Centres providing enterprises with ready-to-use, local, enterprise-grade environment for AI experimentation at STT Singapore 5 facility, offering NVIDIA Blackwell GPUs with 192GB HBM3e VRAM and 400G InfiniBand networking with 14-day free trial.
• SuperX Chat An AI-powered chat platform available through the cloud platform, providing conversational AI capabilities for enterprise users.
• 1.6T Optical Module Solution A high-bandwidth optical interconnect solution for large-scale AI training and inference environments, supporting complete digital signal processing with integrated silicon photonics technology for Ethernet and InfiniBand AI infrastructure architectures.

Quantifiable outcome

• Reduces AI data center deployment timeline from 18-24 months to under 6 months
• +4 more outcomes

Named customers2 records

Segments4 records

Ideal customer profiles3 records

Technology focussed Yes

API detail

Has API

No

API docs

API detail

Core technology

AI maturity

App detail

AI capability7 records

Feature7 records

Partnerships

Nine partnerships are on record, tiered strategic, core and support.

• Kazakhstan InveststrategicStrategic or Co-development Partner · 23 June 2026Kazakhstan's state investment arm and SuperX exploring feasibility of 1GW AI computing park in Kazakhstan. Phased construction: 200MW by 2027, 300MW by 2028, 500MW by 2029. Project intended as cross-border computing node bridging Europe and Asia. Kazakhstan Invest to coordinate on land use, energy supply, and policy frameworks. Discussions ongoing at World Economic Forum Summer Davos.
• ST Telemedia Global Data Centres (STT GDC)coreStrategic or Co-development Partner · 15 April 2026Joint launch of AI Innovation Centre at STT Singapore 5 facility in Tai Seng, Singapore. The center provides enterprise-grade AI compute infrastructure using NVIDIA Blackwell GPUs (192GB HBM3e per GPU) with 400G InfiniBand networking. Offers 14-day free trial enabling enterprises to move from AI experimentation to production-ready scale. Partnership addresses critical AI infrastructure gap across Singapore and Asia, with research showing majority of organizations stalled in early AI development phases.
• Suzhou TFC Optical Communication Co., Ltd. (SZSE: 300394)coreStrategic or Co-development Partner · 11 February 2026Joint venture SuperX Optical Communications Pte. Ltd. established in Singapore, combining SuperX's AI infrastructure expertise with TFC's optical component manufacturing capabilities. Targets global AI optical connectivity markets excluding mainland China, Hong Kong, and Macau. Addresses high-speed data transmission bottlenecks in AI computing clusters through integrated optical interconnect products with plug-and-play deployment capabilities. SuperX leads through board and management control.
• Digital Dynamic Inc., eole Inc., and Woodman Inc.strategicStrategic or Co-development Partner · 4 February 2026Memorandum of Understanding signed to co-develop large-scale AI Data Centers across Japan. Joint task force to pilot AIDC project in Mie Prefecture with initial capacity up to 4MW, serving as reference for potential expansion up to 300MW total capacity. Leverages SuperX's modular architecture for next-generation liquid-cooled GPU infrastructure. Subject to feasibility assessments and regulatory approvals.
• Zhonhen Electric Co., Ltd.coreStrategic or Co-development Partner · 31 October 2025Joint venture SuperX Digital Power Pte. Ltd. launched two flagship 800VDC power solutions: Panama-800VDC for new-build AI data centers and Aurora-800VDC for retrofitting existing facilities. Achieves up to 98.5% system efficiency, reduces power conversion stages from 4-5 to 1-2, and delivers up to 3.6 MW per rack. Solutions support hyperscale AI and accommodate increasing power demands of modern GPU systems.
• Beijing Teamsun Technology Co., Ltd.supportStrategic or Co-development Partner · 30 October 2025Joint venture SuperX Global Service established in Singapore to provide AI infrastructure services globally. Venture enhances SuperX's product lifecycle offering and strengthens market position in AI infrastructure by combining product and service capabilities.
• MicroInference Pte. Ltd.coreStrategic or Co-development Partner · 24 October 2025SuperX acquired majority stake in Singapore-based MicroInference, an NVIDIA Partner Network Solution Provider specializing in AI data center solutions. The investment enhances SuperX's AI ecosystem, strengthens supply chain for NVIDIA servers and networking equipment, and accelerates deployment of AI infrastructure across Asia Pacific. MicroInference operates as a subsidiary within SuperX's joint venture structure.
• Shenzhen Chengtian Weiye Technology Co., Ltd. (SZSE: 300689)coreStrategic or Co-development Partner · 24 October 2025Joint venture SuperX Cooltech Pte. Ltd. established in Singapore to develop liquid cooling solutions for AI data centers. SuperX becomes single largest shareholder. Leverages Chengtian Weiye's R&D and manufacturing capabilities in Micro Channel Liquid Cold Plates, CDU systems, and manifolds. Targets global markets excluding mainland China, Hong Kong SAR, and Macau. Responds to escalating power demands from next-generation GPU architectures exceeding 100kW per rack.
• NVIDIAcoreTechnology or IntegrationNVIDIA is the foundational technology partner powering SuperX's AI infrastructure solutions. SuperX integrates NVIDIA GPUs including HGX B200/B300, RTX PRO 6000 Blackwell GPUs, GB300 NVL72 systems, and ConnectX networking. The partnership involves certified system designs, access to latest GPU architectures, and co-development of rack-scale AI platforms.

Scale indicators9 records

Recent moves2 records

Expansion highlights6 records

Direct peers

• Super Micro Computer: Super Micro is the closest direct peer — a leading provider of GPU-accelerated AI servers (HGX-based systems), liquid-cooled rack solutions, and full-stack data center platforms built on NVIDIA architectures. Both companies target hyperscalers, sovereign AI, and enterprise AI infrastructure deployments with similar product portfolios.
• CoreWeave: CoreWeave is a specialized AI cloud provider offering NVIDIA GPU-accelerated compute on demand, competing directly with SuperX's Denver AI Inference Cloud Hub and AI Cloud platform. Both target enterprise customers needing access to Blackwell-generation GPUs without large upfront capital.

Broad incumbents

• Dell Technologies: Dell's PowerEdge XE9680 and XE8640 servers with NVIDIA HGX H200/B200/GB200 GPUs directly compete with SuperX's AI server line. Dell's broader enterprise infrastructure portfolio, global services organization, and established hyperscaler relationships make it a formidable incumbent in the same customer segments.
• Hewlett Packard Enterprise: HPE's Cray XD and ProLiant Gen11 AI servers, combined with its GreenLake private cloud platform, compete with SuperX's full-stack AI infrastructure offering. HPE's enterprise customer base and Cray supercomputing heritage position it as an incumbent in both AI training and sovereign/HPC deployments.
• Lenovo Group: Lenovo's ThinkSystem SR675 V3 and SR780a V3 with NVIDIA HGX platforms, plus its Neptune liquid cooling technology, make it a direct competitor in AI servers. Lenovo's strong Asia Pacific presence, particularly in Japan and Singapore, creates head-to-head competition with SuperX in its core geographies.
• Inspur Electronic Information Industry: Inspur is one of the world's largest AI server manufacturers with NF5288M6 and other HGX-based systems. While primarily focused on mainland China, Inspur's scale and NVIDIA partnerships make it a relevant benchmark for AI server manufacturing economics and a potential competitor in Asia Pacific.

Emerging players

• Vertiv Holdings: Vertiv provides power, cooling, and IT infrastructure for data centers, directly competing with SuperX Digital Power (800VDC) and SuperX Cooltech (liquid cooling) JVs. Vertiv's Liebert products and global services network are entrenched in hyperscale and colocation deployments.
• Lambda Labs: Lambda provides GPU cloud and on-premise AI workstations/servers (Hyperplane, Lambda Stack), competing with SuperX's AI Cloud and Workstation (XK1140/XK1270) product lines. Both target AI developers and enterprises needing flexible access to NVIDIA-based compute.

Others

• NVIDIA Corporation: NVIDIA is SuperX's foundational technology partner and the supplier of all GPU silicon in its products. NVIDIA's growing direct-to-customer DGX Cloud and reference architecture efforts represent both the demand engine and a competitive threat to AI infrastructure integrators like SuperX.
• Equinix: Equinix is a global colocation and interconnection provider that could host SuperX's modular AI factory deployments or compete in the AI-ready data center space. The AI Innovation Centre partnership with STT GDC (a regional data center operator) indicates SuperX is following a similar ecosystem model.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat5 records

Key risks7 records

Key highlights7 records

Customer concentration

• Xx.com/SUPERX_AI_
• Facebookfacebook.com/people/Super-X-AI-Technology-Limited/61578918040072
• LinkedInlinkedin.com/company/superx-ai

Compliance1 record

Revenue estimate

Valuation estimate

Number of profiles

Profiles6 records

Subsidiaries9 records

Funding overview

Funding rounds2 records

Investors1 record

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

M&A1 record

Investments

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

Frequently asked questions about Super X AI