NeuralMesh and Augmented Memory Grid Integration with NVIDIA STX Increases Token Production by 6.5x in the Same GPU Footprint, Slashing Cost of Inference for AI-Driven Organizations

SAN JOSE, Calif. and CAMPBELL, Calif., March 17, 2026 /PRNewswire/ — From GTC 2026: WEKA, the AI storage and memory systems company, today announced the integration of its NeuralMesh™ software with the NVIDIA STX reference architecture. WEKA’s breakthrough Augmented Memory Grid™ memory extension technology running on NeuralMesh will support NVIDIA STX to bring high-throughput context memory storage to agentic AI factories, making long-context reasoning seamless across sessions, tools, and tasks. Leveraging NVIDIA Vera Rubin NVL72, NVIDIA BlueField-4, and NVIDIA Spectrum-X Ethernet, the NeuralMesh solution based on NVIDIA STX will deliver an estimated increase of 4-10x more tokens per second for context memory while supporting at least 320 GB read and 150 GB write throughput per second for AI workloads, more than double the throughput of conventional AI storage platforms.

WEKA and NVIDIA unlock cost-efficient AI inference at scale

Solving the Inference Cost Problem with Shared KV Cache Infrastructure
Scaling agentic systems, especially for software engineering applications, exposes a hard truth: today’s AI economics are decided at the memory infrastructure layer. Every large-scale inference fleet hits the memory wall: limited high-bandwidth memory (HBM) on the GPU is rapidly exhausted, key-value (KV) cache is evicted, context is lost, and the system is forced to repeat work it already completed. This architectural inefficiency sends inference costs soaring. The answer is a shared KV cache infrastructure that keeps context live across agents, users, and sessions. It eliminates redundant computation, sustains token throughput, and maintains predictable performance. Without shared KV cache infrastructure, every increase in concurrent users and agents becomes a liability — costs rise, experiences degrade, and the inference fleet becomes harder to operate the larger it grows. With STX for context memory, NVIDIA is introducing a blueprint to address these core inference bottlenecks.

Context Memory Storage: The Foundation of Agentic AI Factories
With co-designed WEKA solutions based on NVIDIA STX architecture, AI clouds, enterprises, and AI model builders can deploy the infrastructure foundation they need to run GPUs at peak productivity, sustain high-volume token production, and make large-scale inference more energy and cost-efficient.

Leading AI innovators and cloud providers, such as Firmus, are already transforming their inference economics with Augmented Memory Grid on NeuralMesh.

“Real-world AI doesn’t run in a lab— it has power constraints, cooling limits, and relentless workload demand. Firmus is built for exactly that. Paired with NVIDIA AI infrastructure, WEKA Augmented Memory Grid delivers up to 6.5x higher tokens per second and 4x faster TTFT at scale, proving we can get more performance from the same GPU footprint. With NeuralMesh and Augmented Memory Grid integrated into our NVIDIA-aligned AI Factory and NVIDIA STX reference architecture, we’ll be able to deliver the fastest context memory network for predictable and efficient inference at scale,” said Daniel Kearney, Chief Technology Officer at Firmus.

NeuralMesh and NVIDIA STX: Purpose-Built for Agentic AI
NeuralMesh is WEKA’s intelligent, adaptive storage system built on over 170 patents. It will run across the full-stack STX reference architecture, providing the next-generation storage organizations need to standardize high-performance AI data services and accelerate agentic AI outcomes. WEKA’s Augmented Memory Grid is a purpose-built memory extension layer that pools and persists KV cache outside of GPU memory, keeping long-context sessions stable and concurrency high as inference workloads grow. First unveiled at GTC 2025 and generally available to NeuralMesh customers today, Augmented Memory Grid has been validated with Supermicro on NVIDIA Grace CPUs and BlueField-3 DPUs to deliver numerous benefits that improve AI economics, including:

• Faster User Experiences: Augmented Memory Grid on NeuralMesh delivers up to 4-20x improvement in time-to-first-token, keeping AI agents and applications responsive under real-world load.
• More Revenue from the Same Hardware: Serve 6.5x more tokens per GPU — without adding infrastructure.
• Sustained Performance at Scale: Augmented Memory Grid maintains high KV cache hit rates even as sessions, agents, and context windows grow — preventing the performance cliff that hits DRAM-only architectures.
• GPU-Native Efficiency: BlueField-4 integration offloads the storage data path from the CPU, keeping GPUs fully productive and eliminating I/O bottlenecks.

“With coding LLMs advancing, we’re seeing unprecedented adoption of Agentic AI use cases for software engineering, where productivity increases by 100-1000x. As coding assistants make repeated calls against largely unchanged codebases and prompts, WEKA’s Augmented Memory Grid reuses cached context instead of forcing redundant prefill, even as context windows grow to incredible lengths. This provides a major boost in response times and greatly increases the number of concurrent users running on the same infrastructure,” said Liran Zvibel, co-founder and CEO at WEKA. “WEKA first identified this need for context memory storage more than a year ago and launched Augmented Memory Grid at GTC 2025. Now, NVIDIA STX opens the door to organizations running their storage and memory extension infrastructure on state-of-the-art NVIDIA Vera Rubin architecture, including NVIDIA BlueField-4 and NVIDIA Spectrum-X Ethernet. Running Augmented Memory Grid on NeuralMesh for NVIDIA STX delivers extreme performance and efficiency that translates directly to game-changing AI economics.”

Availability

WEKA’s Augmented Memory Grid is commercially available with NeuralMesh today.

Organizations that don’t address the memory wall today will find it harder and more expensive to scale tomorrow. As agentic workloads grow and context windows expand, DRAM-only architectures face a compounding cost problem: each additional concurrent user or session increases recomputation overhead, GPU idle time, and operational cost. The organizations that architect for persistent KV cache now will have a structural cost and performance advantage over those that wait.

For more information about NeuralMesh, visit: weka.io/NeuralMesh.
For more information about Augmented Memory Grid, visit: weka.io/augmented-memory-grid.

Organizations can learn more at weka.io/nvidia or visit WEKA at GTC 2026, booth #1034.

About WEKA
WEKA is transforming how organizations build, run, and scale AI workflows with NeuralMesh™ by WEKA®, its intelligent, adaptive mesh storage system. Unlike traditional data infrastructure, which becomes slower and more fragile as workloads expand, NeuralMesh becomes faster, stronger, and more efficient as it scales, dynamically adapting to AI environments to provide a flexible foundation for enterprise AI and agentic AI innovation. Trusted by 30% of the Fortune 50, NeuralMesh helps leading enterprises, AI cloud providers, and AI builders optimize GPUs, scale AI faster, and reduce innovation costs. Learn more at www.weka.io or connect with us on LinkedIn and X.

WEKA and the W logo are registered trademarks of WekaIO, Inc. Other trade names herein may be trademarks of their respective owners.

WEKA_v1_Logo_new

SEOUL, South Korea and KATOWICE, Poland, March 17, 2026 /PRNewswire/ — LX Pantos, a leading global logistics company (Global CEO: Lee Yong-ho), is accelerating its global expansion through strategic partnerships and infrastructure investments.

On March 16, LX Pantos announced that a consortium including the Korea Overseas Infrastructure & Urban Development Corporation (KIND) and the PIS No. 2 Fund, a policy fund under Korea’s Ministry of Land, Infrastructure and Transport, has acquired a newly built logistics complex in Katowice, southern Poland, in a transaction valued at approximately USD 148 million.

The Katowice Logistics Center consists of five buildings with a combined total floor area of 109,000㎡. The complex is being developed and brought into operation in phases, with full completion scheduled for the first half of 2026. Several global and local companies have already secured leases.

Through the Katowice Logistics Center, LX Pantos plans to establish an Eastern European logistics hub and expand services for key industries such as automotive parts, consumer goods, and home appliances. The facility is also expected to serve as a forward logistics base to support reconstruction-driven demand in Ukraine once rebuilding begins.

Beyond Europe, the company is actively expanding its global operations in key strategic markets, including the United States and Asia. In the United States, LX Pantos established Boxlinks LLC, a joint venture with Ocean Network Express (ONE), and acquired a large-scale logistics center in Georgia with a site area of 304,769 m², equivalent to approximately 43 soccer fields, to strengthen its intermodal operations. In Northeast Asia, the company formed FutureLinks with Sinotrans to expand China–Korea multimodal transport and tap the fast-growing Sea & Air market driven by China-origin e-commerce cargo.

“LX Pantos operates subsidiaries in 40 countries and more than 380 global networks, delivering integrated sea, air, rail, and contract logistics solutions,” said Lee Yong-ho, Global CEO of LX Pantos. “Against this backdrop, the Katowice Logistics Center represents a strategic milestone in strengthening our presence in the European logistics market.”

• About LX Pantos

Established in 1977, LX Pantos is a leading global logistics provider headquartered in Korea. It delivers comprehensive logistics solutions across sea, air, rail, and contract logistics through a worldwide network spanning more than 40 countries.

New NeuralMesh AI Data Platform Closes the Gap Between AI Proof-of-Concept and Profitable Production, Delivering Scalable Business Intelligence and Faster AI Outcomes with NVIDIA

SAN JOSE, Calif. and CAMPBELL, Calif., March 17, 2026 /PRNewswire/ — From GTC 2026: WEKA, the AI storage and memory systems company, today announced general availability of its enterprise-ready NeuralMesh™ AI Data Platform (AIDP), which delivers composable, high-performance infrastructure optimized for AI Factory deployments. Based on NVIDIA AI Data Platform reference design, the solution is an end-to-end system that accelerates the delivery of AI-ready data to AI factories. The result: AI project timelines speed up from months to minutes, empowering organizations to deliver production-scale agentic AI applications using best-in-class technologies across their ecosystem.

WEKA and NVIDIA accelerate enterprise-ready AI factories

Leveraging NeuralMesh’s uniquely adaptive architecture, the solution addresses the most persistent obstacle in enterprise AI: organizations can demonstrate AI concepts work in proof-of-concept (POC) but consistently struggle to reach production scale.

Built on more than 170 patents and over a decade of AI-native storage innovation, a foundation no competing storage platform can replicate, NeuralMesh is the only solution that gets faster and more resilient as AI environments scale to exabytes and beyond. As AI Factory data infrastructure becomes a critical layer in enterprise AI architecture, NeuralMesh is helping customers to close the gap between POC and production deployments today. Customers running NeuralMesh with Augmented Memory Grid™ can achieve 6.5x more tokens per GPU for inference workloads, reflecting the compounding advantage of a purpose-built architecture over retrofitted infrastructure.

“Enterprises are now deploying AI Factories internally, driving a major shift to inference throughout the ecosystem. These companies require rapid AI outcomes and need turnkey solutions that come with the enterprise table-stakes of reliability, security, and optimal price-performance and cost-effectiveness,” said Liran Zvibel, cofounder & CEO at WEKA. “WEKA’s NeuralMesh AIDP gives organizations everything they need to run always-on AI factories: extreme storage performance and the flexible architecture required to operationalize AI at production scale. Whether an organization is just beginning its AI journey or running full-stack NVIDIA deployments, NeuralMesh AIDP scales seamlessly as they grow.”

“The deployment of agentic AI in production demands a new focus on managing the continuous, coherent flow of data and inference context,” said Jason Hardy, vice president, storage technologies at NVIDIA. “By leveraging the NVIDIA AI Data Platform, solutions like WEKA’s NeuralMesh AIDP deliver the persistent context tier necessary for stable and high-scale agentic inference.”

One System, Every AI Workload: Delivering End-to-End AI Factories

AI factories provide enterprises with purpose-built production systems designed to operate AI at scale, but they demand storage capabilities that extend beyond where data sits to actively support context and continuous data movement. NeuralMesh, WEKA’s intelligent, adaptive storage system, delivers the continuous data-loop performance that AI factory workloads demand.

Out-of-the-Box AI Applications Designed to Accelerate Business Outcomes

NeuralMesh AIDP enables enterprises and AI cloud providers to unify AI operations from retrieval to inference on a single, ready-to-deploy platform. With pre-integrated hardware and software options from NVIDIA (including NVIDIA RTX 6000 PRO Server Edition GPUs and the newly announced NVIDIA RTX 4500 PRO Server Edition GPUs) alongside Red Hat, Spectro Cloud and Supermicro, organizations can eliminate months of AI integration work.

The platform provides a simplified solution that allows teams to focus on intelligence output rather than managing underlying infrastructure. It delivers ready-to-use pipelines for a spectrum of business use cases that work across verticals, including: Semantic Search, Video Search & Summarization (VSS), AlphaFold for drug discovery, AIQ/Agentic RAG and more.  

These AI applications are already being used by enterprise and research customers to drive outcomes across high-priority sectors:

• Health & Life Sciences: Identify patient subgroups across multiple studies and accelerate discovery in data-intensive workflows such as cryo-EM.
• Financial Services: Get early market signal detection as data lands and institutionalize knowledge access into a shared, secure resource.
• Public Sector: Detect potential threats based on context and meaning, not keywords, and automate evidence synthesis across sources to improve decision-making cycles.
• Physical AI & Robotics: Shorten the loop from real-world data capture to retrained model deployment, improving fleet performance, reliability, and time to market.

“The missing piece in production AI isn’t reasoning models or compute power. It’s having an efficient platform that unifies the AI Factory pipeline and makes it truly scalable,” said Shimon Ben-David, CTO at WEKA. “The NeuralMesh AIDP was designed to close AI’s production and profitability gap, taking enterprise experiments to full-scale operations and making AI economically viable for everything from next-generation agents to healthcare applications.”

Supporting Partner & Customer Quotes

“Getting AI to production requires more than technology— it requires consistency and control. By using the NeuralMesh AI Data Platform with Red Hat AI Enterprise, based on Red Hat OpenShift, organizations can run data-intensive AI pipelines across on-premises and cloud environments at the scale that enterprise production demands, without sacrificing governance or security,” said Ryan King, vice president, AI and Infrastructure Partners at Red Hat.

“The real challenge in AI is no longer training models. It is running them reliably in production, at scale, with predictable performance and cost. That’s where most AI initiatives stall. The NeuralMesh AI Data Platform integrates with our AI Acceleration Cloud, Neysa Velocis, to solve that problem directly. It gives teams a way to run AI workloads as dependable systems, without carrying the operational burden of stitching together complex infrastructure,” said Anindya Das, cofounder and CTO at Neysa.

Availability
The NeuralMesh AI Data Platform solution is available now, delivered as an appliance-style system. Organizations can learn more at weka.io/nvidia or visit WEKA at GTC 2026, booth #1034 for a demo.

For more information on the NeuralMesh AIDP:

• Blog: weka.io/blog/ai-ml/neuralmesh-aidp-built-to-operationalize-AI-at-enterprise-scale/
• Solution Webpage: weka.io/product/neuralmesh-aidp/

About WEKA

WEKA is transforming how organizations build, run, and scale AI workflows with NeuralMesh™ by WEKA®, its intelligent, adaptive mesh storage system. Unlike traditional data infrastructure, which becomes slower and more fragile as workloads expand, NeuralMesh becomes faster, stronger, and more efficient as it scales, dynamically adapting to AI environments to provide a flexible foundation for enterprise AI and agentic AI innovation. Trusted by 30% of the Fortune 50, NeuralMesh helps leading enterprises, AI cloud providers, and AI builders optimize GPUs, scale AI faster, and reduce innovation costs. Learn more at www.weka.io or connect with us on LinkedIn and X.

WEKA and the W logo are registered trademarks of WekaIO, Inc. Other trade names herein may be trademarks of their respective owners.

WEKA: The Foundation for Enterprise AI

The global #1 in wet and dry hard floor cleaning goes big for the Amazon Big Smile Sale

SYDNEY, March 17, 2026 /PRNewswire/ — Tineco, the global #1 leader in wet and dry hard floor cleaning, is bringing a smile to the faces of house-proud Australians this week, with up to 56% off its wet and dry floor cleaners and stick vacuums range as part of the Amazon Big Smile Sale.

The start of autumn, wet weather and the extra dirt and mud coming through the house means it’s a great time to invest in convenient, deep and hygienic cleaning at home. Running from today through to Sunday, March 22, the Amazon Big Smile Sale will give Australians access to Tineco’s leading cleaning features, including its renowned HyperSteam, HyperStretch and FlashDry self-cleaning technologies, all at a fraction of the normal recommended retail prices.

Tineco’s model line-up for this year’s Amazon Big Smile Sale deals combine design and performance, with hard to beat savings:

• FLOOR ONE Stretch S6 – RRP $899 reduced to $399 – 56% off
• FLOOR ONE i7 Stretch Steam – RRP $999 reduced to $599 – 40% off
• FLOOR ONE S7 Artist – RRP $899 reduced to $579 – 36% off
• FLOOR ONE Switch S7 Stretch – RRP $1199 reduced to $799 – 33% off
• PURE ONE Station 5 –  RRP $799 reduced to $559 – 30% off

A key feature that sets Tineco wet and dry floor cleaners apart is the brand’s proprietary MHCBS (Maintain Hygiene Clean Brush System) Technology, which involves a floating scraper continuously washing and rinsing the roller brush with fresh water while also keeping it free from tangles, to ensure a spotless finish and hygienic floor after cleaning. This makes a significant difference in the effectiveness of the floor clean, with only clean water used at all times of the cleaning process.

Tineco’s ongoing commitment to provide feature-rich and value-for-money cleaning solutions for consumers has led the company to be recognised as the #1 global leader in the household wet & dry floor cleaner category* for the fourth consecutive year by Euromonitor International, the world’s leading independent provider of strategic market research.

Tineco has also been the #1 wet & dry floor cleaner brand on Amazon in Australia, United States, Canada, France, Italy, and Japan for three years running, with the brand working hard to deliver deep discounts once again as part of this year’s Big Smile Sale.

Chris Loong of Tineco says: “Millions of consumers around the world put their trust in the Tineco brand to deliver deep, effective floor cleaning solutions for their home. The Amazon Big Smile Sale is the ideal opportunity for Australians to secure significant savings on our range of wet and dry floor cleaners and stick vacuums, exclusively available through Amazon Australia.”

“Whether you live in an apartment or have a busy household with children and pets, our range of wet and dry floor cleaners integrate cleaning technologies such as steam power and HyperStretch Technology with 180° lay-flat design to ensure that dust, dirt and messes have nowhere to hide.”

Tineco’s FLOOR ONE Stretch S6 (RRP $899, now $399 – 56% OFF) cleans up wet and dry messes in one step and offers Dual-sided Edge Cleaning to get close up to skirtings and walls. Its 45° swivel design helps easily manoeuvre around furniture such as chairs and table legs and combined with a lay-flat design that reduces to just 13cm to get under low furniture, the Stretch S6 has daily spot cleaning covered. The inclusion of Tineco’s signature FlashDry self-cleaning system that uses 70° fresh water and air to clean and dry from the pipes to the rollers means hands-free maintenance and an appliance that is always at the ready for its next task.

Featuring Tineco’s premium HyperSteam Technology, the FLOOR ONE i7 Stretch Steam (RRP $999, now $599 – 40% OFF) uses super-heated steam (up to 140℃ and reaching the floor at no less than 99℃) to dissolve stubborn grease, stains and sticky residue from floor surfaces. The 180° lay-flat design effortlessly reaches low areas, enabling the cleaning of hidden dust at heights as low as 13 cm. An 80-min run time allows for whole-house cleaning on a single charge and FlashDry self-cleaning makes post-use clean up quick and easy.

Boasting an elegant design, Tineco’s FLOOR ONE S7 Artist (RRP $899, now $579 – 36% OFF) subtly blends into the aesthetic of modern homes while retaining high performance, effective deep cleaning with powerful 22kPa suction and a 50-min run time. Its ultra-slim lay-flat design means the appliance compresses to just 12.85cm to fit under beds and sofas, and DualBlock Anti-Tangle scrapers prevent hair tangling and clogging – ideal for homes with fur-friends sharing living areas. FlashDry self-cleaning also features in this model, giving consumers every reason to smile thanks to a value-packed deal.

Tineco’s FLOOR ONE Switch S7 Stretch (RRP $1,199, now $799 – 33% OFF) is a 5-in-1 multi-function cleaner that features a SwitchPro Motor to switch between floor washer and vacuum for whole-house cleaning on the go. DualBlock Anti-Tangle design and ZeroTangle Brush design targets hair messes and pet fur with ease and prevents tangling. An upgraded FlashDry self-cleaning system uses fresh water heated to 85°C to effectively dissolve stains from the pipe to the brush roller after use and the 85°C hot air effectively dries every part of the machine.

Tineco’s PURE ONE Station 5 (RRP $799, now $559 – 30% OFF) automatically self-cleans and recharges when returned to its station, saving users time and hassle of after-use cleaning. This model features full-path self-cleaning meaning the brush, tube, filter and dustbin are thoroughly cleaned after each use. Its Station Dura-cyclone System features multiple cyclones and a self-cleaning HEPA filter to maintain strong, consistent suction power over time, and a unique brush head design, combined with a wider tube and a larger connection port inside the vacuum, allows the Station5 to easily pick up large debris without clogging.

Tineco’s Amazon Big Smile Sale deals are available at www.amazon.com.au from 16^th March to 22^nd March 2026. To learn more about Tineco’s Amazon range of intelligent stick vacuums, floor washers, and carpet cleaners, visit https://www.amazon.com.au/tineco

ABOUT TINECO
Tineco (“tin-co”) was founded in 1998 with its first product launch as a vacuum cleaner and, in 2019, pioneered the first-ever smart vacuum. Today, the brand has evolved into a global leader in intelligent appliances spanning floor care, kitchen, and personal care categories. With a growing user base of over 23 million households and availability in approximately 30 countries worldwide, Tineco remains committed to its brand vision of making life easier through smart technology and continuous innovation. For more information, visit https://au.tineco.com/  

• Nscale Acquires the Monarch Compute Campus — America’s First State-Certified AI Microgrid with a Potential of up to 8GW+ of Onsite Powered Microgrid

SAN JOSE, Calif., March 17, 2026 /PRNewswire/ — Nscale has signed an agreement to acquire American Intelligence & Power Corporation (“AIPCorp”), sponsored by Fidelis New Energy and 8090 Industries, including the Monarch Compute Campus in West Virginia — with plans to build one of the world’s largest AI Factories.

The acquisition secures the Monarch Compute Campus, a site with up to 2,250 acres in Mason County, West Virginia, and the United States’ first state-certified AI microgrid with a power runway scalable to over 8 gigawatts. It also establishes Nscale Energy & Power, a new global division of Nscale headquartered in Houston, Texas. The site has access to long-term, low-cost, onsite powered microgrid to support multi-gigawatt expansion over the long-term, a key bottleneck in the buildout of in-demand, cutting-edge AI compute.

The West Virginia campus is the first state-certified utility grid purpose-built for AI workloads in the United States. Initial power capacity of 2 gigawatts is expected to be online by the first half of 2028 with an expansion to approximately 8 gigawatts planned for 2031.

“Nscale is a global company, and the US is the world’s largest AI infrastructure market. AI infrastructure needs to be built where demand is, and right now a significant share of that demand is in the United States,” said Josh Payne, CEO of Nscale. “Monarch allows us to meet that demand. The acquisition builds on our existing US footprint and reflects the pace at which we are scaling to serve customers around the world.”

Nscale Energy & Power

Nscale Energy & Power will be led by veteran energy executive Daniel Shapiro as Chief

Power Officer, alongside Bengt Jarlsjo as Deputy Chief Power Officer. The full AIPCorp and Fidelis New Energy organization — leadership and staff — joins Nscale as part of the acquisition. AIPCorp’s team brings nearly a decade of experience and continuity in AI infrastructure development and will support Nscale’s development of the Monarch Compute Campus and Nscale’s future global development and expansion operations.

“The vision for AIPCorp was always to build the vertically integrated power-and-compute model that this combination achieves instantly. This combination — plus Nscale’s disciplined balance sheet — delivers exactly what the high-growth AI market demands: a debottlenecked, integrated neocloud platform built to scale,” said Daniel Shapiro, Chief Power Officer.

Morgan Stanley & Co. LLC acted as financial advisor, and Latham & Watkins served as legal counsel, each to Nscale. Greenhill, a Mizuho affiliate, acted as exclusive financial advisor to AIPCorp, and Vinson & Elkins LLP served as legal counsel to Fidelis New Energy.

About Nscale

Nscale is building the global hyperscaler engineered for AI infrastructure. Through vertically integrated AI solutions and modular, first-principles datacenter design across Europe and North America, Nscale delivers the compute foundation for enterprise AI training, fine-tuning, and inference at scale.

Media Contact: press@nscale.com 

• Nscale Acquires the Monarch Compute Campus — America’s First State-Certified AI Microgrid with a potential of up to 8GW+ of onsite powered microgrid
• Nscale Signs Letter of Intent with Microsoft for up to 1.35GW of AI Compute using NVIDIA Vera Rubin NVL72 GPUs
• Establishes Nscale Energy & Power, Positions Nscale as Global Flagship Deployment Partner for NVIDIA Vera Rubin Architecture, NVIDIA DSX AI Factory

SAN JOSE, Calif., March 17, 2026 /PRNewswire/ — Nscale has signed a letter of intent with Microsoft to provide 1.35 gigawatts of AI compute capacity, setting up the West Virginia Monarch AI campus as a global flagship deployment of NVIDIA’s next-generation Vera Rubin GPUs as Vera Rubin NVL72 systems engineered with the NVIDIA Vera Rubin DSX AI Factory reference design.

Nscale today also announced the acquisition of American Intelligence & Power Corporation (“AIPCorp”), sponsored by Fidelis New Energy and 8090 Industries, which includes the Monarch Compute Campus, a site with up to 2,250 acres in Mason County, West Virginia, and the United States’ first state-certified AI microgrid with a power runway scalable to over eight gigawatts.

Under the collaboration with Microsoft, Nscale will construct and operate advanced AI data center infrastructure to host this large-scale GPU deployment based on the latest generation NVIDIA Vera Rubin NVL72 GPUs and future technologies. The deployment will be delivered across multiple tranches beginning in late 2027, creating one of the largest dedicated AI compute installations in the world.

The planned infrastructure will operate under a long-term framework that includes an initial multi-year compute services term alongside a long-term data centre lease structure, reflecting the scale and strategic importance of the facility. This positions the campus to be a cornerstone facility for next-generation AI training and inference capacity in the United States.

“This collaboration with Microsoft marks a pivotal milestone both for Nscale and the development of the Monarch Campus,” said Josh Payne, CEO of Nscale. “By integrating our specialized AI infrastructure with Microsoft’s global platform, we are creating a foundation for innovation that can scale alongside the most ambitious AI models in the world”.

This acquisition and collaboration comes at a crucial moment. AI is forecast to drive a sharp increase in global data center demand, but estimates suggest existing supply is constrained by the pace at which power and new capacity can be brought online. McKinsey estimates AI-related data center capacity demand could reach 156 GW by 2030. The Monarch Compute Campus will build on Nscale’s current capacity of over 1GW.

The expansion capabilities of this site — which has the potential for a total power draw of 8GW — will allow Nscale to develop power capacity rapidly while supporting one of the largest announced AI compute deployments in the market today.

The campus is also expected to provide high-speed fiber connectivity to some of the nation’s largest AI hubs. Its close proximity to major centers of AI and cloud infrastructure, including Ashburn and Chicago, will offer customers low latency and minimal delays for AI workloads.

“Microsoft’s datacenter approach is to build the best global infrastructure informed by near-term and long-term demand,” said Jon Tinter, President, Business Development and Ventures at Microsoft. “Our investments blend owned datacenters, leased facilities, and strategic collaborations. This collaboration with Nscale and NVIDIA is an important step to deliver meaningful AI innovation to our customers.”

“AI is becoming essential infrastructure for every industry,” said Nico Caprez, Vice President, Global AI Infrastructure Growth, NVIDIA. “With this large-scale NVIDIA DSX AI Factory Blueprint, Nscale is building the infrastructure required to produce intelligence at industrial scale and power the next wave of global innovation.”

Power collaboration with Caterpillar

Through a strategic collaboration with Caterpillar, Nscale will deploy Caterpillar G3500 series natural gas generator sets at sufficient scale to achieve two gigawatts of power generation by the first half of 2028, powering the NVIDIA Vera Rubin DSX AI Factory reference design.

“This collaboration reflects Caterpillar and our dealers’ continued focus on supporting customers that require primary, continuous-duty power at scale through our broad energy solutions portfolio,” said Melissa Busen, senior vice president of Electric Power, Caterpillar. “Projects like Monarch demonstrate how Caterpillar’s natural gas generation platforms are being deployed as core infrastructure for data centers and other power intensive applications where reliability, speed of deployment, and lifecycle performance are critical.”

The G3500 series units provide Nscale with a proven, rapid-deployment power solution that collapses traditional infrastructure timelines and accelerates the path from site to live compute.

Providing positive community impact

Nscale is working closely with state and local officials and community partners to ensure the development delivers lasting value for the region. The campus is being designed with local and environmental resources in mind. Power will be generated on-site as the facility operates independently of the local grid, eliminating the burden on existing utility customers and protecting ratepayers’ bills. The microgrid is also designed to enable future tie-in to the grid for export of power back to the grid.

Nscale is also pursuing carbon sequestration to offset emissions, with access to significant sequestration capacity in West Virginia. Even at full 8GW capacity, the campus will use a high-efficiency design that consumes less water with no impact on municipal water supply or residential users.

About Nscale

Nscale is building the global hyperscaler engineered for AI infrastructure. Through vertically integrated AI solutions and modular, first-principles datacenter design across Europe and North America, Nscale delivers the compute foundation for enterprise AI training, fine-tuning, and inference at scale.

Media Contact: press@nscale.com