Automating the Edge with Robotics

Arm breaks new ground with new Neoverse compute subsystems

Categories Edge Computing News  |  Hardware
Arm breaks new ground with new Neoverse compute subsystems

Arm recently announced that it will release two new Neoverse compute subsystems (CSS) built on a brand new third-generation of its Neoverse IP.

The Neoverse cloud-to-edge microarchitecture launched in October 2018 and features 64-bit ARM processor cores. N-Series is for data centers, E-Series is for edge computing and N-Series excels in high-performance computing, the company says..

In 2023, Arm introduced Neoverse CSS to facilitate the integration of Arm Neoverse-based technology into new compute solutions. This technology allowed partners to create customizable silicon more affordably and quickly than previous IP solutions.

Unlike AMD and Intel, Arm licenses core designs to companies that sell its chips. Other organizations customize these designs with elements like storage, memory and PCLe controllers.

The new Neoverse CSS N3 by Arm prioritizes power efficiency, delivering a 20% increase in core performance per watt compared to the CSS N2. The CSS N3 is also built on the company’s new Neoverse N3 IP platform and, in turn, brings Arm v9.2 features to the N-Series, the company says.

Arm also notes that it extends the CSS program to V-Series cores with Neoverse CSS V3 for an over 50% socket performance boost compared to CSS N2. V3 scales up to 128 cores per socket and supports modern IO standards and high-speed memory.

Dermot O’Driscoll, the VP of product solutions, infrastructure line of business, at Arm, says that when it comes to performance, the most significant improvement is the gain for AI data analytics workloads, with the V3 having an 84 percent performance uplift over the V2. The N3 has a 196 percent performance uplift when measured on simulation.

The V3 provides hardware support for the Arm Confidential Computing architecture. The V3 and N3 cores feature private L3 cache sizes.

Arm also collaborates with its ecosystem partners to address chiplet physical layer compatibility challenges. By introducing a new chiplet system architecture, they aim to create an interoperable ecosystem with 15 partners to maximize chiplet technology potential.

Mohamed Awad, the SVP and general manager of Arm’s infrastructure line of business, highlights that the newest Neoverse platforms will underpin future products and services created by partners. He continues that this move comes as AI workloads increasingly permeate various computing domains.

“From the largest data centers to networks, servers, and all aspects of the computing infrastructure, Neoverse V3 and N3 provide both the computing performance and efficiency needed to underpin this transformation and our compute subsystems supported by the Arm Total Design ecosystem allow for faster time to market and more investment in innovation than ever before,” states Awad.

The Arm Total Design ecosystem, launched four months ago, now includes 20 members, incorporating EDA, IP providers and silicon design partners like Broadcom, Faraday, Intel Foundry Services, Samsung and TSMC.

Read more:

Arm announces a power efficient Cortex-M processor with Helium technology

Arm’s new chip designs and Nvidia-powered boards from AverMedia

Article Topics

 |   |   | 

Comments

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Featured Edge Computing Company

Edge Ecosystem Videos

Automating the Edge

“Automating

Deploying AI Models at the Edge

“Deploying

Latest News