The UK-based Arm, a company that manufactures systems-on-chip ( SoCs ), is used by some of the biggest tech companies in the world, including Google’s parent company Alphabet and others, without ever producing any hardware of its own, though that reportedly will change this year.
And you’d think it might want to simply stay raking in the money given its record-setting past third of$ 1.24 billion in total income.
Finger wants a piece of the action because some of its customers are offering record profits of their own by offering AI graphics processing units that incorporate Leg technology, and Shoulder sees how quickly AI has taken off in the business.
The business changed its focus from providing part Internet to a platform-first business now with a new item naming strategy.
It’s about demonstrating to customers that we have much more to give than just device and technology designs. In an exclusive meeting with VentureBeat over Zoom yesterday, Leg chief marketing officer Ami Badani stated in an exclusive interview with VentureBeat that” we have a full habitat that can help them size AI and do so at lower cost with greater efficiency.
Indeed, Arm’s history of producing lower-power chips than the competition ( cough cough, Intel ) has proven to be very effective in providing the foundation for power-hungry AI training and inference jobs, as CEO Rene Haas pointed out to tech news outlet Next Platform back in February.
According to his remarks in that content, today’s data center uses 460 terawatt hours of electricity annually, which is expected to doubled by the end of the decade and may account for 25 % of the world’s energy consumption unless more Arm power-saving chip styles and their accompanying tailored software and firmware are employed in the network for these locations.
A significant shift from IP to platforms
Arm is restructuring its offerings to accommodate complete compute platforms as AI workloads increase in complexity and power requirements.
Partner building AI-capable chips can now integrate more quickly, scale more effectively, and reduce complexity thanks to these platforms.
Arm is reviving its previous naming conventions and introducing new product families organized by market to reflect this change:
- Neoverse for infrastructure
- Niva for PCs
- Lumex for mobile
- Zena for automobile
- Orbis for edge AI and IoT
The Mali brand will continue to represent the integrated GPU offerings in these new platforms.
Arm is updating its product numbering system in addition to the renaming. The IP identifiers for Ultra, Premium, Pro, Nano, and Pico will now match platform generations and performance tiers. This structure aims to make the roadmap more accessible to both users and developers.
supported by positive outcomes
Arm’s strong Q4 fiscal year 2025 ( ended March 31 ), where the company crossed the$ 1 billion mark for the first time, led to the rebranding.
Total revenue hit$ 1.24 billion, up 34 % year-over-year, driven by both record licensing revenue ($ 634 million, up 53 % ) and royalty revenue ($ 607 million, up 18 % ).
Notably, this increase in royalty rates was fueled in part by the adoption of Arm Compute Subsystems ( CSS) across smartphones, cloud infrastructure, and edge AI as a result of the Armv9 architecture’s increased use.
The mobile market was a standout: Arm’s smartphone royalty revenue increased by about 30 % while global smartphone shipments increased by less than 2 %.
Additionally, the business also signed its first automotive CSS agreement with a leading global EV manufacturer, allowing it to gain more ground in the high-growth automotive sector.
Arm hasn’t yet given the precise name of the EV manufacturer, but Badani told VentureBeat that it sees automotive as a major growth market in addition to AI model providers and cloud hyperscalers like Google and Amazon.
The CMO told VentureBeat that” we’re looking at automotive as a major growth area and we think AI and other advances like self-driving are going to be standard, which our designs are perfect for.”
As a result, cloud providers like AWS, Google Cloud, and Microsoft Azure kept deploying Arm-based silicon to run AI workloads, strengthening Arm’s growing influence in data center compute.
establishing a new platform ecosystem with vertically integrated products and software
Arm is expanding its hardware platforms with more software and ecosystem support.
Users can use Arm’s architecture to optimize code using its free GitHub Copilot extension, which is now available for all developers.
Arm’s Kleidi AI software layer has over 8 billion installed devices, surpassing the 22 million that it currently employs.
The rebranding is seen by Arm’s leadership as a natural progression in its long-term strategy. The company’s goal is to meet the growing demand for energy-efficient AI compute from device to data center by offering vertically integrated platforms with performance and naming clarity.
In Arm’s blog post, Haas stated that Arm’s compute platforms are the foundation of a future where AI is ubiquitous, and Arm is on a mission to build that foundation at scale.
What does it mean for decision-makers who use AI and data?
This strategic repositioning will likely change the way technical decision-makers in all roles in AI, data, and security approach their day-to-day tasks and future plans.
The more organized platform structure facilitates the selection of compute architectures that are best suited for AI workloads for those managing large language model lifecycles.
With predefined compute systems like Neoverse or Lumex, model deployment timelines are tightening as the bar for efficiency rises, reducing the overhead needed to evaluate raw IP blocks and allowing faster execution in iterative development cycles.
The modularity and performance tiering of Arm’s new architecture could aid in the standardization of AI pipelines for engineers who manage them across environments.
It provides a practical framework for managing resource-intensive training tasks in the cloud or adjusting compute capabilities to fluctuating workload requirements.
These engineers may find more clarity in mapping their orchestration logic to predefined Arm platform tiers because they frequently juggle system uptime and cost-performance tradeoffs.
Leaders of data infrastructure who are in charge of upholding high-throughput pipelines and ensuring data integrity may also gain.
The system-level integration and naming change indicate a deeper commitment from Arm to support scalable designs that work well with AI-enabled pipelines.
The compute subsystems may also shorten the time to market for custom silicon that supports next-generation data platforms, which is crucial for teams that operate under tight budgets and limited engineering resources.
Security experts, in contrast, are likely to observe effects on how system-level compatibility and embedded security features develop within these platforms.
Security teams can more easily plan for and enforce end-to-end protections because Arm aims to offer a consistent architecture across edge and cloud, especially when integrating AI workloads that demand both performance and strict access controls.
Enterprise architects and engineers are now sent a message that Arm is no longer just a component provider; it offers full-stack foundations for how AI systems are built and scaled.
English 
Romanian 