Semiconductor design company Arm has introduced two compute platforms, one for data centre silicon aimed at agentic AI and one for mobile devices covering both agentic workloads and neural graphics.
For the data centre, the company has launched Neoverse Compute Subsystems (CSS) N4. Arm describes it as its most configurable CSS to date and its fastest route from a compute subsystem to finished silicon.
Go deeper with GlobalData
Access deeper industry intelligence
Experience unmatched clarity with a single platform that combines unique data, AI, and human expertise.
Arm claims the design delivers up to twice the performance, up to 1.25 times the performance per watt and up to 1.75 times the memory bandwidth of Neoverse CSS N3.
Neoverse CSS is intended for partners building silicon tailored to a particular system architecture or function, from bespoke compute platforms to data processing units (DPUs) and networking designs.
Arm says the pre-integrated approach lowers engineering effort, integration risk and time to silicon. Its Arm Total Design ecosystem supports earlier IP validation and software development, a collaborative model the company is now extending to physical AI.
Customers can use CSS N4 to build silicon optimised for maximum throughput efficiency, according to the company.
Arm AGI CPU, by contrast, offers production-ready compute designed for highly responsive agentic AI. Both share a common Neoverse platform and software ecosystem.
Arm cloud AI executive vice-president Mohamed Awad wrote: “There is no one-size-fits-all approach to AI infrastructure, so Arm gives customers flexibility to choose what best fits their workloads.”
Separately, Arm has introduced CSS for Mobile 2. The platform combines the new Arm Mali G2-Ultra NX graphics processing unit (GPU), the Arm C2 CPU cluster with SME2, enhanced system IP, physical implementations and a developer-ready software ecosystem.
Arm edge AI executive vice-president Chris Bergey wrote: “Mobile devices need a compute platform built for this new era, bringing CPUs, GPUs and system IP together in a way that developers can easily utilise and apply to next-generation AI workloads.”
The Arm Mali G2-Ultra NX is the first Mali GPU with dedicated neural accelerators, placing neural and traditional graphics processing in the same pipeline. It also adds a new execution engine and next-generation ray tracing unit.
Used with Arm Neural Technology, it is claimed to deliver up to four times higher performance per watt for neural graphics.
The accompanying Arm C2 CPU cluster pairs the C2-Ultra, Arm’s most powerful mobile CPU, with efficiency-focused C2-Pro cores and two SME2 units. The doubled SME2 capability will enable a 70% speedup on the latest small language models, according to the company.
Against C1-Ultra, C2-Ultra offers up to 1.7 times higher AI performance and 15% higher single-thread performance, while using up to 38% less power at the same performance.
