SAN JOSE, Calif., Oct. 14, 2026 /PRNewswire/ -- PEGATRON, a globally recognized provider of comprehensive AI infrastructure and server solutions, will showcase its capabilities in developing and manufacturing next-generation AI infrastructure at OCP Global Summit 2026. Under the theme "Scaling Innovation for the AI Era," Pegatron will demonstrate how advanced AI platforms can be transformed into production-ready systems, racks, and clusters for rapid deployment in data centers.
As AI infrastructure scales from individual servers to large-scale clusters, successful deployment requires more than compute performance. It requires system-level engineering, liquid cooling, power architecture, open standards, manufacturing readiness, and rigorous validation. Pegatron brings these capabilities together through OCP-aligned architectures, including ORv3 busbar, DC-MHS, and OCP 3.0 network adapters, enabling modular and scalable infrastructure for modern data centers. Its OCP-compliant platforms span AI rack solutions, accelerated computing, compute, and storage infrastructure.
At the center of the showcase is the NVIDIA Vera Rubin NVL72, Pegatron's RA4803-72N3, a rack-scale AI platform integrating 72 NVIDIA Rubin GPUs, 36 NVIDIA Vera CPUs, 6th generation NVIDIA NVLink, and fully liquid-cooled rack-level integration. The platform highlights Pegatron's ability to engineer and manufacture complex rack-scale AI solutions and migrate from design to cluster deployment.
Extending beyond accelerated computing, Pegatron will showcase a NVIDIA Vera CPU liquid-cooled system and NVIDIA Groq 3 LPX, addressing the evolving infrastructure requirements of agentic AI. NVIDIA Vera, the CPU for agents, is optimized to execute tool calls, sandboxed code, and reinforcement-learning loops between model steps. Deployed at rack scale, NVIDIA Vera CPU provides dense capacity to run alongside GPU infrastructure, supporting efficient execution, orchestration, and data processing workloads in AI factories. NVIDIA Groq 3 LPX further extends the NVIDIA Vera Rubin platform for highly interactive, latency-sensitive inference.
Pegatron will also feature the AS210-2T1-8H3, a 2U fully liquid-cooled NVIDIA HGX Rubin NVL8 system combining dual Intel® Xeon® 6 processors, AMD EPYC™ 9006 processors, or NVIDIA Vera CPUs with an eight NVIDIA Rubin GPU baseboard. The platform demonstrates high-density system engineering, fully liquid-cooled architecture, and rack-level integration. Its rack-scale configurations include the RA4800-64H3, supporting eight systems and up to 64 GPUs, and the RA4800-72H3, supporting nine systems and up to 72 GPUs in a 48U liquid-cooled rack.
The portfolio also includes AI-ready compute and storage infrastructure powered by AMD EPYC™ 9006 Series processors. Pegatron will showcase the SS201-1A1, a 2U single-socket storage server supporting up to 32 E3.S NVMe SSDs for high-throughput AI data pipelines, alongside the GS203-2A2 and GS205-2A2E 2U dual-socket platforms for enterprise, cloud, HPC, and data-intensive workloads. With up to 256 cores and 512 threads, high memory bandwidth, and PCIe® Gen 6 connectivity, AMD EPYC 9006 processors provide a high-density compute foundation across compute and storage applications. These platforms incorporate OCP 3.0, DC-MHS, and flexible power architectures, including 48VDC busbar configurations for OCP racks.
From digital engineering to factory deployment, Pegatron is extending its AI infrastructure capabilities through NVIDIA DSX AI factory platform-aligned workflows. At its facilities, digital-twin simulation and optimization are connected with physical manufacturing, automated production, thermal validation, and system to rack level run-in testing, helping translate engineering insights into production-ready AI infrastructure.
"A key measure of AI infrastructure is how efficiently it can move from engineering to deployment," said Gary Cheng, President and Co-CEO of Pegatron. "Pegatron combines deep system design expertise with vertically integrated manufacturing, OCP-aligned architecture, and rack-level validation before deployment to rapidly build and deploy AI servers, clusters, and data center infrastructure at scale. By connecting digital-twin simulation, thermal validation, automated manufacturing, and rack-level testing, we help customers move faster from AI infrastructure design to production deployment."
Through its integrated capabilities, Pegatron is enabling AI infrastructure to move beyond individual systems toward production-ready AI clusters and data centers—bringing advanced AI infrastructure from concept and simulation to validated, scalable deployment.
For more information, please visit PEGATRON SVR website and follow us on LinkedIn and YouTube.
Website: https://svr.pegatroncorp.com
LinkedIn: https://tw.linkedin.com/showcase/pegatron-svr/
YouTube: https://www.youtube.com/@PEGATRONSVR
About PEGATRON
PEGATRON Corporation (hereafter referred to as "PEGATRON"), with abundant product development experience and vertically integrated manufacturing, is committed to providing clients with innovative design, systematic production, and manufacturing services to comprehensively and efficiently satisfy all our customers' needs.
Drawing on accumulated experience in server design, manufacturing, and deployment, Pegatron focuses on developing a variety of state-of-the-art servers including liquid-cooled/air-cooled server solutions based on x86 and ARM architectures, racks, and AI clusters, that meet the requirements of present and future Cloud Service Providers' data centers, as well as enterprise-grade data centers.
Pegatron Corporation Website: https://www.pegatroncorp.com/
** This press release is distributed by PR Newswire through automated distribution system, for which the client assumes full responsibility. **
PEGATRON Brings AI Infrastructure from Design to Deployment at OCP Global Summit 2026
Brightray expands into the world's largest data center market with its first U.S. project letter of intent, targeting delivery within one year with Brightray prefabricated data center solution to support next-generation AI compute deployment
SINGAPORE, Oct. 14, 2026 /PRNewswire/ -- Brightray Science Inc. (together with its subsidiaries, "Brightray"), a majority-owned subsidiary of BUUU Group Limited (Nasdaq: BUUU), today announced that its wholly owned subsidiary, Brightray Veridian Pte. Ltd., entered into a letter of intent (the "LOI") on October 6, 2026 with a U.S.-based mission critical infrastructure partner (the "Partner") for a planned 10MW AI data center project in Indiana, United States. The LOI is Brightray's first project letter of intent signed in the United States and marks an important step in its expansion into the world's largest data center market.
The project site is located in Indiana, United States. Under the LOI, the Partner intends to engage Brightray to deploy a 10MW prefabricated data center solution, with implementation subject to the parties reaching final agreement on technical specifications, commercial terms, the project schedule and definitive contract documentation. Brightray intends to act as the supplier for the prefabricated data center scope agreed by the parties, providing related design, manufacturing, integration, delivery and other agreed services. The principal cooperation arrangements under the LOI are not legally binding and do not create any obligation on either party to purchase, supply or make payment. Under the LOI, the Partner intends to issue a purchase order to Brightray by November 15, 2026. The LOI will expire 12 months after it was signed unless replaced earlier by a definitive agreement. There can be no assurance that the parties will enter into a definitive contract, that the Partner will issue a purchase order or that the project will proceed as planned.
Project Overview
The project is intended to use Brightray prefabricated data center solution, with planned power capacity of 10MW. To meet the U.S. market's pressing need for speed of delivery, Brightray intends to complete the pre-installation and integration of the data hall modules and related functional systems at the manufacturing stage, so that on-site work is concentrated on module placement, interface connections and integrated commissioning. The single-storey layout will be designed around the equipment configuration, operation and maintenance access and site conditions, providing space and facility support for the deployment of high-density AI equipment.
The project is intended to be compatible with the NVIDIA GB300 computing platform, with power distribution, cooling and data hall layout planned around the requirements of high-density AI equipment deployment. The project is planned to be designed on the basis of Tier III standards, with a target design annual power usage effectiveness ("PUE") of below 1.2, aiming to balance high-density computing capacity, operational reliability and energy efficiency.
By carrying out factory prefabrication in parallel with on-site foundation and ancillary works, Brightray plans to shorten the project delivery period to within one year. The one-year target covers the overall project schedule in the United States, including design finalization, permitting, international shipping, on-site foundation and ancillary works, installation and commissioning, and not only the delivery of the prefabricated modules. The starting point of the delivery period and the implementation schedule will be specified in the subsequent definitive contract. The final design, technical specifications and delivery arrangements of the project remain subject to factors including the parties' final agreement, site and power conditions, required approvals and equipment supply.
Entering the World's Largest Data Center Market, Capturing AI Build-Out Demand Through Delivery Capability
The United States is the world's largest data center market and a major center of global AI infrastructure investment and construction. According to JLL's 2026 Global Data Center Outlook, the Americas account for approximately 50% of global data center capacity, with the United States accounting for approximately 90% of capacity in the Americas. The scale of this market makes the United States an important market for Brightray's international expansion and provides broad scope for construction solutions that can accelerate the delivery of computing infrastructure.
Strong market demand comes with significant delivery constraints. CBRE's U.S. Real Estate Market Outlook 2026 notes that power supply and infrastructure conditions continue to affect data center construction schedules and that traditional 12-to-18-month timelines for sub-50-MW buildings no longer apply. Against this backdrop, speed of delivery has become an important consideration for customers in selecting construction solutions. By carrying out factory prefabrication in parallel with on-site works, Brightray intends to shorten the delivery period for this project to within one year; the starting point and delivery conditions will be specified in the definitive contract.
The LOI gives Brightray the opportunity to carry out its first project in the U.S. market. For Brightray, its significance lies not only in a new 10MW project opportunity, but also in bringing the prefabricated delivery model to the world's largest data center market and building local project execution and partnership experience. If the technical solution, commercial arrangements and definitive contract are finalized, the project is expected to serve as a reference for Brightray's future business expansion in the United States.
Brightray Prefabricated Data Center Solution Supports Rapid Delivery of AI Infrastructure in the United States
Mr. Bin Wang, Founder of Brightray, said, "Entering the U.S. market is an important step in Brightray's international development. The U.S. has enormous data center demand, and customers face real pressure to turn their AI investments into usable computing power as quickly as possible. Our entry point is delivery: with Brightray prefabricated data center solution, we move more of the work into the factory and run manufacturing in parallel with on-site works, with the aim of shortening the delivery period for this project to within one year."
Mr. Wang added, "We look forward to working further with our partner on the technical solution and implementation arrangements and to building delivery experience in the U.S. market through this 10MW project. Addressing the power, cooling and space requirements of next-generation AI equipment, we aim to bring the efficiency advantages of our prefabricated data center solution to the project and lay the foundation for the further expansion of our U.S. business."
About Brightray
Brightray is an AI infrastructure industrialization company and a 60%-owned subsidiary of BUUU Group Limited (Nasdaq: BUUU). Brightray accelerates data center deployment through prefabricated, modular and standardized solutions, with services spanning design, manufacturing, delivery and full lifecycle support. Its portfolio of prefabricated data center solutions, allows Brightray to match the appropriate solution to each customer's computing needs, site and building conditions, project scale and delivery schedule.
For more information, please visit www.brightraydc.com.
About BUUU Group Limited
BUUU Group Limited (Nasdaq: BUUU) has established industrialized AI data center delivery as its core growth business, operated through Brightray Science Inc., in which BUUU acquired a controlling interest in September 2026. BUUU also provides MICE (meetings, incentives, conferences and exhibitions) solutions spanning event management and stage production to public institutions, agencies, real estate corporations and established brands.
Forward-Looking Statements
This press release contains forward-looking statements within the meaning of Section 27A of the Securities Act of 1933, as amended, Section 21E of the Securities Exchange Act of 1934, as amended, and the Private Securities Litigation Reform Act of 1995. Forward-looking statements can be identified by words such as "may," "will," "expect," "anticipate," "aim," "intend," "plan," "believe," "target" or similar expressions. Forward-looking statements include statements regarding the progress of the cooperation contemplated by the LOI, the proposed purchase order and the timing of its issuance, the execution and terms of definitive agreements, the project's design and capacity, GB300 compatibility, design PUE, the Tier III design target, the delivery period, and Brightray's business plans and prospects in the U.S. market.
These statements are based on current expectations and assumptions and involve risks and uncertainties, many of which are beyond the control of BUUU and Brightray, that could cause actual results to differ materially from those expressed in the forward-looking statements, including, but not limited to: the non-binding nature of the principal cooperation arrangements under the LOI and the possibility that the parties may not enter into definitive agreements or proceed with the transaction; the possibility that the Partner may not issue a purchase order when expected or at all; the parties' ability to agree on final technical specifications, commercial terms, project scope and schedule; changes to the project's design, capacity or schedule; land and site conditions; power supply and grid connection timing; required planning, building, power and other approvals; project financing; the Partner's financial condition and ability to perform; the progress of equipment procurement, manufacturing, transportation and on-site construction; differences between actual operating conditions and design assumptions; the integration of Brightray with BUUU; competition in the data center industry and demand for AI computing infrastructure; changes in applicable laws and regulations and in trade, tariff and export control measures; and other risks described in the documents BUUU files with or furnishes to the U.S. Securities and Exchange Commission. Forward-looking statements speak only as of the date of this press release, and, except as required by applicable law, BUUU undertakes no obligation to update them. NVIDIA and GB300 are trademarks of NVIDIA Corporation. Neither BUUU nor Brightray has any agreement with NVIDIA Corporation, and references to third-party products imply no affiliation or endorsement.
** This press release is distributed by PR Newswire through automated distribution system, for which the client assumes full responsibility. **
Brightray Signs Letter of Intent for Planned 10MW AI Data Center Project in the United States
