The Advanced Capabilities of Arm Neoverse: Looking Ahead

Optimized for performance, power, and cost.

Meet the Arm Neoverse Family

Arm Neoverse CPUs are designed to handle a wide range of cloud-native workloads at world-class levels of performance, efficiency and compute density.

Arm Neoverse differs from traditional infrastructure CPUs in a number of critical ways:

Arm Neoverse processors provide a full core to each single thread. This leads to consistent performance, better scalability, improved economics, and improved security – the key attributes demanded by cloud customers.
Arm Neoverse-based CPUs are optimized for different workloads and can be combined with Marvell’s own IP to build CPUs, DPUs, 5G RAN accelerators and other solutions.
The Arm Corelink Coherent Mesh Network family (CMN) links individual cores to each other, as well as to I/O, memory, or other accelerators and system IPs. Arm’s latest CMN mesh supports new integration techniques, including CXL and chiplets (using UCIe or AMBA CHI) that make new Infrastructure solutions possible. These technologies put full control over performance, power, area, cost, and schedule back in the hands of silicon designers.

The Marvell OCTEON 10 DPU is based on the Neoverse N2 core. Learn more about Neoverse N2 here.

Discover Arm Neoverse

Arm Neoverse V-series CPUs are designed for maximum per-thread performance and are suitable for server ‘host’ use cases, such as cloud and HPC, where results are time-sensitive, or customers are paying by the hour.
Arm Neoverse N-series is built for efficient scalability. N-series cores are effective for “always-on” workloads that need to deliver consistent, high performance but in power- or thermally constrained environments. The N-series is also ideal for ‘scale-out’ CPU solutions packing a maximum core count into a less constrained TPD.
Arm Neoverse E-series is designed for high throughput-based devices such as edge servers, IoT gateways, and smart equipment, where data throughput with energy and space efficiency is paramount.

Performance and Sustainability

The V and N series in particular deliver performance within the constraints of sustainability required in demanding situations, such as scalability/performance for 5G. They efficiently move the network and compute power to the edge and are designed to handle the explosion in data and the need for high-performance analytics.

Greater overall performance. Fugaku, the world’s second fastest supercomputer, runs on the Arm-based Fujitsu A64FX processor. Overall, Fugaku contains more than 159,000 chips running more than 8 million Arm CPUs.
Greater performance per watt. Customers report that Arm Neoverse-based CPUs perform 2x to 3.5x more operations per watt than traditional processors in datacenters and 5G networks.
Greater performance per rack. More cores mean more operations within the same physical footprint for a better ROI. By 2030, widespread use of Arm Neoverse could increase the world's datacenter capacity by 25%.
Greater performance per dollar. Arm Neoverse delivers 20% to 40% uplifts in performance, and 20% to 50% reductions in cloud costs, often at the same time.
Greater security. A multiple core/single thread design physically isolates operations from each other, reducing the 'noisy neighbor' problem and improving security.