ARM Announces New Neoverse N1 and E1 Platforms for Scalable Cloud-to-Edge Infrastructure

Recently, ARM announced the brand‑new Neoverse N1 and E1 platforms, optimized with a 7 nm process to provide high scalability, massive throughput and strong performance, injecting new momentum for 5G and future IoT development and driving the next wave of infrastructure platform transformation.

In October of last year, Arm released the Neoverse family to build a cloud‑to‑edge infrastructure for a world of a trillion connected devices. Since then the ecosystem has grown rapidly, with partners such as Huawei, HPE (which used Arm for the world’s first Top‑500 supercomputer based on Arm) and AWS Graviton.

These smart devices will inevitably change data usage patterns in the coming years, prompting adjustments in storage, processing, and information management. Arm is shifting from a video‑distribution model (e.g., Netflix, YouTube) to a model that consumes, manages, and processes data generated at the edge.

Arm Neoverse N1: Accelerating the Transition to Scalable Cloud‑to‑Edge Infrastructure

With the Neoverse N1 platform, Arm lifts next‑generation infrastructure‑grade compute performance to the highest levels in the industry, exceeding original expectations: integer performance improves up to 60% and key cloud workloads see up to 2.5× gains. To help partners fully exploit N1’s efficiency, performance and space‑saving advantages, Arm has optimized the leading 7 nm process.

The N1 platform is built from the ground up and, beyond raw compute, offers infrastructure‑grade features such as server‑class virtualization, advanced RAS support, power and performance management, and system‑wide analytics. It also includes a uniform mesh interconnect, industry‑leading power efficiency, and a highly integrated compact design that scales from 4 to 128 cores. This scalability lets partners add accelerators or custom logic to their own SoCs, creating flexible compute solutions and delivering lower total‑cost‑of‑ownership (TCO) for infrastructure customers.

Arm Neoverse E1: Empowering Infrastructure to Meet Next‑Generation Throughput Demands

The surge in data not only demands higher compute performance but also efficient processing, analysis, filtering and storage. As the world transitions to 5G, network infrastructure is undergoing a major transformation.

The Neoverse E1’s unique design enables a smooth transition from 4G to a more scalable 5G infrastructure, meeting diverse compute requirements. Compared with previous Arm solutions, E1 delivers more than 2.7× higher throughput, over 2.4× better efficiency, and more than 2× higher compute performance. It supports a wide range of devices—from sub‑35 W base stations to routers handling hundreds of gigabytes—by providing scalable throughput from edge to core.

Thanks to a rich software ecosystem, support for traditional network protocol stacks, and leading open‑source SDN solutions, the entire network stack can fully leverage the advantages of the Neoverse E1 platform.

Server Turning Point?

In recent months the Arm server ecosystem has been exciting. At last year’s Hot Chips conference, Fujitsu unveiled the A64FX HPC processor, marking the company’s shift from SPARC to ARMv8 and delivering the first chip to implement Arm’s new Scalable Vector Extension (SVE).

Cavium’s ThunderX2 achieved a remarkable performance leap, becoming one of the first processors to compete with Intel and AMD, and together with partners such as GIGABYTE it offers full‑system server solutions based on the new SoC.

Huawei recently released the Kunpeng 920 server chip, which is expected to become the highest‑performance Arm‑based server CPU in the industry.

The common thread among these three products is that each represents a vendor’s effort to build a custom micro‑architecture under an Arm‑v8 licence. This raises the question: what are Arm’s own plans for the server and infrastructure market? For those closely watching, the newly announced Neoverse series is not surprising, as Arm first revealed the brand and roadmap back in October.

Big Debate: Which Will Dominate the Server Market, x86 or ARM?

The launch of Neoverse N1 and E1 has attracted the attention of Linux creator Linus Torvalds, who remains skeptical about the ARM server market despite acknowledging recent progress. A commenter named Michael S argued that Linus does not fully understand the benefits of cross‑development for most developers, which is precisely where ARM excels. Linus countered that ARM‑compatible processors will never dominate the server market, citing stability concerns if everyone adopts cross‑development.

Redis creator Salvatore Sanfilippo offered a different view, stating that he is moving Redis to ARM as its primary architecture; some low‑level Redis code runs only on ARM and all tests pass with excellent stability, so he disagrees with Linus’s stance.

Many developers feel Linus’s view was historically correct but overlook the possibilities brought by rapid technological change. Recent news shows Apple will likely ship ARM chips in Macs next year, and Microsoft is adding more ARM support to Windows 10. Overall, the future for the ARM platform looks bright.

Recommended Reading:

NVMe over Fabric Protocol and I/O Flow

OLC NAND Flash, Achieving 256 Voltage States?

Warm Tip:

Please scan the QR code to follow the public account and click the original link to obtain more technical material summaries.