What YTL AI Cloud does

YTL AI Cloud Sdn Bhd is a Malaysian AI infrastructure-as-a-service provider founded in 2024 as a wholly owned subsidiary of YTL Power International Berhad, an international multi-utility conglomerate operating across power generation, water, telecommunications, and digital technologies. The company deploys NVIDIA GB200 NVL72 GPU clusters within a purpose-built, 1,640-acre green data center campus in Johor, Malaysia — located less than 30km from Singapore and powered by an on-site 500MW solar facility with direct-to-chip liquid cooling supporting rack densities of 100kW or more. At full Phase 1 scale the platform exceeds 300 exaflops of AI compute, making the YTL AI Supercomputer one of the fastest globally. Initial buildings are designed for more than 80MW of IT load, with Phase 1 scaling to 500MW in a single campus.

The commercial stack integrates NVIDIA Spectrum-X Ethernet networking and a proprietary Cloud Orchestration Platform (COP) that dynamically manages compute, network topology, and high-performance storage to deliver deterministic, bare-metal performance. The company's offerings are structured around two motions: pay-as-you-go On Demand GPU access (RM 68.80/hour for NVIDIA GB200 compute with 192GB HBM3e, plus RM 0.60/hour for YTL AI Object Storage) for developers and researchers, and a Cloud Agreement Framework (CFA) of multi-year committed-usage contracts distributed through appointed Managed Service Providers for enterprise customers. YTL AI Cloud targets enterprise AI developers, ML researchers, public sector agencies pursuing sovereign AI workloads, and academic institutions across Southeast Asia. The company holds a broad certification portfolio (SOC 2, ISO 27001/27017/27018/27701/22301/14001/20000-1, CSA STAR L2, HIPAA, GDPR, FedRAMP Moderate, NIST 800-53/800-171, GxP) and was one of the first cloud providers in Asia to achieve NVIDIA Exemplar Cloud status in May 2026, validated across a 512-GPU environment on frontier model workloads.

Name

YTL AI Cloud

Legal name

YTL AI Cloud Sdn Bhd

Website

https://ytlaicloud.com

Company type

Private

Founded year

2024

Operating status

Operating

Short description

YTL AI Cloud is a Malaysian AI infrastructure provider founded in 2024 as a YTL Power International subsidiary, offering NVIDIA GB200 NVL72 GPU clusters from a Johor campus to enterprise developers, researchers, and sovereign AI customers across Southeast Asia.

Ownership category

akta.pro rank

Product category

Cloud Infrastructure

NAICS

Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services (5182)

SIC

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

akta.pro primary industry

AI Server Systems & HGX/Accelerator Platforms (HDAAAAAB)

akta.pro secondary industries

AI Accelerators (GPUs/TPUs/NPUs/ASICs) (HDAAAAAA), General-Purpose Compute (VMs & Bare Metal IaaS) (HDABAAAA), AI Compute Virtualization & Scheduling (GPU virtualization, cluster schedulers) (HDAAAAAG)

Keywords

• GPU cloud computing
• AI infrastructure services
• Cloud GPU rental
• Sovereign AI cloud
• AI supercomputing platform

Headquarters

HQ city

Kulai

HQ country

Malaysia

HQ region

Asia

Offices3 records

Markets served

GTM type

B2B

Offering type

Software

Cost components

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

Revenue model

1. GPU Cloud Computing - On Demand: Usage-based pricing for GPU compute resources and storage. Customers pay per hour for GPU usage (NVIDIA GB200 at RM 68.80/hour) and storage (RM 0.60/hour). Provides immediate access to GPU infrastructure without upfront capital investment.
2. Cloud Agreement Framework (CFA): Subscription-based committed usage contracts through appointed Managed Service Providers. Offers discounts over on-demand prices for customers who commit to longer-term usage via enterprise agreements.

Pricing tiers

Go-to-market motion2 records

Distribution channels3 records

Marketing channels4 records

Core offering

YTL AI Cloud provides massive-scale GPU cloud computing infrastructure built on NVIDIA GB200 NVL72 architecture, accessible as on-demand GPU compute and enterprise subscriptions through the Cloud Agreement Framework. The platform is orchestrated by a proprietary Cloud Orchestration Platform (COP) and backed by YTL AI Object Storage, enabling AI training, inference, and Sovereign AI workloads for enterprise and developer customers.

Product overview

YTL AI Cloud is a specialized AI infrastructure-as-a-service provider offering a unified GPU cloud computing platform. The core offering consists of On Demand GPU compute resources (NVIDIA GB200 with 192GB HBM3e) and YTL AI Object Storage, supported by the proprietary Cloud Orchestration Platform (COP). The platform is anchored by the YTL AI Supercomputer — one of the world's fastest supercomputers exceeding 300 exaflops of AI compute — built on NVIDIA Grace Blackwell DGX Cloud architecture. YTL AI Cloud operates from the YTL Green Data Center Campus in Johor, Malaysia, drawing power from on-site solar facilities. YTL AI Labs contributes the MalayMMLU benchmark for Bahasa Melayu language model evaluation. The company also offers GPU cloud pricing calculator for cost estimation.

Differentiator

Problem solved

Functional benefit

Products and services

• On Demand GPU Pay-per-hour GPU compute service using NVIDIA GB200 GPUs with 192GB HBM3e memory and 8.0 TB/s bandwidth, priced at RM 68.80/hour. Provides immediate provisioning of GPU resources for AI training, inference, and graphics-intensive workloads without long-term commitments.
• YTL AI Object Storage Object storage service providing 1.5TB + 1.9TB warm storage capacity at RM 0.60/hour for AI workloads, including model training datasets and inference outputs.
• Cloud Agreement Framework (CFA) Enterprise Subscriptions Subscription-based committed-usage enterprise contracts offering discounts over on-demand prices for customers who commit to longer-term usage. Distributed through appointed Managed Service Providers with multi-year contract options.
• YTL AI Supercomputer Compute Access Access to the YTL AI Supercomputer — one of the world's most advanced AI supercomputers exceeding 300 exaflops of AI compute, built on NVIDIA Grace Blackwell DGX Cloud with Quantum InfiniBand interconnect, housed in the YTL Green Data Center Campus in Johor, Malaysia.

Quantifiable outcome

• 30x increase in real-time trillion-parameter LLM inference performance
• +6 more outcomes

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 capability6 records

Feature4 records

Partnerships

Three partnerships are on record, tiered core/flagship.

• Universiti Malaya (UM)core/flagshipStrategic or Co-development Partner · 24 September 2024Collaboration to develop the first MalayMMLU (Massive Multi-task Language Understanding) benchmark for evaluating large language models in Bahasa Melayu. Accepted at EMNLP 2024 conference. The MalayMMLU consists of over 24,000 questions covering the K-12 spectrum across 22 subjects from the Malaysian education curriculum. YTL AI Labs and Universiti Malaya publicly released the dataset and evaluation code to enable broader AI community contributions.
• NVIDIAcore/flagshipTechnology or Integration · 18 March 2024Strategic collaboration with NVIDIA to provide AI developers with the world's most advanced Model Training and Inference compute infrastructure. YTL is among the first companies to adopt NVIDIA GB200 NVL72, deploying one of the world's most advanced supercomputers on NVIDIA Grace Blackwell-powered DGX Cloud. The platform acts as a single GPU with 1.4 exaflops of AI performance and 30TB of fast memory. YTL AI Cloud achieved NVIDIA Exemplar Cloud status in May 2026, one of the first in Asia.
• YTL Power International BerhadOthers · 18 March 2024YTL AI Cloud is a specialized provider of massive-scale GPU-based accelerated computing infrastructure founded in 2024. It is a wholly owned subsidiary of YTL Power International Berhad, an international multi-utility owner and operator active across power generation, water and sewerage, telecommunications, and digital technologies.

Scale indicators6 records

Recent moves6 records

Expansion highlights6 records

Direct peers

• CoreWeave: Specialized GPU cloud provider offering NVIDIA-accelerated compute for AI training and inference. Closest comparable in business model, target customers (AI labs and enterprises), and NVIDIA-centric stack.
• Lambda Labs: GPU cloud and deep-learning workstation provider serving AI researchers and enterprises. Comparable on NVIDIA GPU rental, AI/ML customer base, and on-demand + reserved pricing models.
• Crusoe Energy: Vertically integrated AI cloud operator building large-scale NVIDIA GPU clusters. Comparable in hyperscale GPU infrastructure focus, sustainability-aligned energy sourcing, and enterprise AI workload targeting.
• Together AI: AI cloud platform providing GPU compute for training and inference of large models. Comparable in serving frontier AI developers, supporting multi-node NVIDIA clusters, and offering both API and raw GPU access.
• Vultr: Cloud provider offering bare-metal and IaaS services including GPU instances. Comparable on flexible hourly GPU compute pricing, self-serve onboarding, and broad developer customer base.
• RunPod: GPU cloud platform targeting AI/ML developers with on-demand and reserved instances. Comparable in NVIDIA GPU focus, developer/AI-researcher GTM, and competitive hourly pricing.
• Paperspace (DigitalOcean): GPU cloud (acquired by DigitalOcean) targeting AI developers and startups. Comparable on self-serve GPU provisioning, AI/ML workload focus, and entry-level developer pricing.

Broad incumbents

• NVIDIA DGX Cloud: NVIDIA's own AI cloud service delivered with partners including YTL. Comparable as the upstream benchmark for validated AI training/inference on Blackwell, but competes indirectly as NVIDIA's flagship cloud offering.
• Oracle Cloud Infrastructure (OCI): Hyperscaler offering bare-metal GPU instances including H100/Blackwell clusters. Comparable as large-scale NVIDIA GPU IaaS provider for enterprise AI workloads.
• Amazon Web Services (EC2 P5 / Blackwell instances): Global hyperscaler offering GPU-accelerated EC2 instances for AI training and inference. Comparable as broad incumbent with NVIDIA Blackwell capacity, mature managed services, and global enterprise reach.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat4 records

Key risks6 records

Key highlights7 records

Customer concentration

Compliance16 records

Revenue estimate

Valuation estimate

Number of profiles

Profiles2 records

Funding overview

Funding rounds

Investors

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 YTL AI Cloud