THE APEX TIMES
Larsen & Toubro’s Vyoma.AI plans an NVIDIA B300 “AI Factory” at its Chennai data centre
The engineering and technology services group says its Vyoma.AI business will deploy NVIDIA B300 systems to support Together AI’s cloud platform for large-scale training and inference.
Larsen & Toubro (L&T) is moving to build what it describes as an NVIDIA B300 “AI Factory” at its Chennai data centre, according to a report citing developments tied to its artificial intelligence platform. The initiative is intended to help run large-scale artificial intelligence workloads, including both training (the process of teaching models using large data sets) and inference (the process of using trained models to produce outputs for users and applications).
The plan centers on deploying NVIDIA B300 systems at the Chennai site, the report says. NVIDIA positions its data center accelerators as engines for running AI workloads, particularly for workloads that require high-performance compute for both model development and production use.
The same report links the capacity build-out to Together AI’s cloud platform, saying L&T’s deployment is meant to support Together AI customers that need access to compute for training and inference at scale. In practice, this kind of arrangement typically reflects demand from AI developers and enterprise teams that want infrastructure without building and operating their own full data centre stacks.
For NVIDIA, B300 is part of the company’s broader push to supply the underlying compute for the rapid expansion of generative AI services worldwide. The report does not provide technical specifications, but an “AI Factory” concept generally implies a purpose-built deployment of multiple AI processing units, networking, and supporting software designed to run sustained model workloads.
L&T’s is positioned in the report as the execution vehicle for the build. L&T has long operated in engineering and infrastructure, and data centre delivery is a natural extension of that capability for customers needing compute capacity in specific geographies.
Still, several commercially important details were not disclosed in the report. It does not specify the total number of NVIDIA B300 units to be installed, the expected timeline to completion, the power and cooling capacity of the Chennai deployment, or any financial terms tied to the arrangement with Together AI.
The report also does not state whether the Chennai “AI Factory” is intended to be exclusively dedicated to Together AI or available more broadly through ’s services. Without disclosures on utilization targets and pricing, it is difficult to gauge how quickly the asset would translate into revenue visibility.
Looking ahead, investors and customers will likely focus on whether L&T and its partners provide further information on build milestones, delivery schedules, and capacity ramp. For NVIDIA, additional announcements tied to named deployments can serve as indicators of demand for its latest data center AI systems, but the near-term impact will depend on disclosed scale and timing.
Why It Matters
- The move highlights continued capital spending by major infrastructure players to expand AI compute capacity in India.
- If executed as described, the Chennai “AI Factory” could help increase access to large-scale training and inference capacity for developers using Together AI’s cloud services.
- For NVIDIA, named deployments of newer data center systems can announcement ongoing demand, though this report does not quantify the scale of the order.
- The lack of disclosed commercial terms means investors will need further updates to assess revenue implications and timing.
Key Facts
- Larsen & Toubro’s plans to construct an “NVIDIA B300 AI Factory” at its Chennai data centre.
- The deployment is described as supporting Together AI’s cloud platform.
- The capacity is intended for large-scale AI training and inference workloads.
- The report ties the plan to NVIDIA’s B300 data center AI systems.
- The announcement does not provide the number of systems, timeline, or financial terms.
Technology Related
Alphabet and Microsoft trade leadership cues as investors focus on who can scale AI compute fastest
A fresh market debate is centered on capital intensity and execution speed as Alphabet and Microsoft vie to supply the growing demand for artificial intelligence workloads.
Boost Run locks in long-term GPU compute expansion lease with 10X Capital
The Nasdaq-listed NVIDIA Preferred Cloud Partner said it finalized a long-term lease agreement tied to expanding GPU capacity, a move aimed at scaling cloud compute services aligned with NVIDIA’s reference architectures.
AI’s next semiconductor bottleneck is memory, and one market note argues investors are looking past Micron and SanDisk
A Yahoo Finance commentary on Aug. 14 said memory is becoming a limiting factor across AI systems, prompting renewed attention to the memory supply chain beyond the best-known names.
Alphabet discloses SpaceX stake now valued near $94 billion, roughly 100-times its original investment
The Google parent said its stake in SpaceX was worth about $94 billion as of the end of June, highlighting how an early bet has compounded over more than a decade.
Chamath Palihapitiya points to Palantir as a potential winner in a shift toward “model-agnostic” AI over the next three years
In a bullish take reported by Yahoo Finance, the venture investor argued that as AI model costs converge, platforms able to work across providers could gain an advantage, with Palantir singled out as a beneficiary of that transition.
LG and NVIDIA Sign Memorandum to Deepen Collaboration in Robotics, AI Factories and Mobility
LG Corp and NVIDIA (NASDAQ: NVDA) have agreed to expand a strategic partnership through a memorandum of understanding focused on robotics, artificial intelligence factory systems, and next-generation mobility technologies, according to a market report published Thursday.
Apple is reportedly building a China-specific AI model with Alibaba’s help, Reuters says
Reuters reported that Apple has trained a large language model tailored for the Chinese market, using assistance from Alibaba. Apple has not publicly detailed the project’s scope or timeline.
Nvidia ties up with Larsen & Toubro on an India AI factory, betting on capacity and local rollout
A new contract announced via Yahoo Finance points to Nvidia expanding its AI infrastructure footprint in India through a major buildout led by Larsen & Toubro, with implications for supply chains and customer access.
Zuckerberg’s “personal agent” vision spotlights AI assistant race, with Meta at the center
Mark Zuckerberg’s latest remarks, as reported by Yahoo Finance, framed a future where people use highly capable personal agents. The comments, while light on specifics, underscore how fast the industry is moving from chatbots to always-on helpers.
Salesforce faces shifting sentiment in latest investor letter as AI roadmap questions resurface
Parnassus Investments’ second-quarter 2026 investor letter highlights how the “AI opportunities shift” can change the appeal of Salesforce’s customer relationship software, even as the fund reports a strong quarterly return.