FEX 2607 Optimizing For Yet-To-Be-Released ARM 256-bit SVE2 Hardware

The CPU's growing role in agentic AI infrastructure

Arm-sponsored content arguing that CPUs play a critical but underappreciated role in agentic AI infrastructure. While accelerators handle model performance, CPUs act as the control plane — managing data movement, workload scheduling, and secure isolation. Arm's Neoverse platform underpins custom silicon from AWS (Graviton), Google (Axion), Microsoft (Azure Cobalt), and NVIDIA (Grace Hopper/Blackwell), all reflecting a shift toward purpose-built Arm-based processors in cloud and AI datacenters. The piece introduces the Arm AGI CPU, built with Meta, targeting rack-level density for agentic AI deployments.

LineShine Is Fastest Supercomputer At Over 2 Exaflops

Siêu máy tính LineShine của Trung Quốc đạt 2,188 exaflops, dẫn đầu danh sách TOP500 nhờ CPU LX2 ARMv9 (304 nhân/core) tích hợp SVE/SME, nhưng tụt xuống hạng 4 ở benchmark HPL-MxP khi các siêu máy tính dùng GPU/NPU như El Capitan (Mỹ) vẫn chiếm ưu thế.

Lập trình viên chuyên về siêu máy tính và parallel computing sẽ tìm hiểu cách tối ưu hóa hiệu suất trên nhiều lõi ARM với SVE/SME để áp dụng vào các giải pháp hiệu năng cao của riêng họ.

How NVIDIA GB10 CPU Performance Compares To Vera

A performance comparison between NVIDIA's GB10 chip (featuring 10 Cortex-X925 and 10 Cortex-A725 cores, 128GB LPDDR5x, 140W TDP) and NVIDIA's Vera CPU (88 custom Olympus cores, 176 threads, 450W TDP). Benchmarks cover single-threaded, multi-threaded, and per-core multi-threaded performance using the Dell Pro Max GB10 and Ubuntu with Linux 6.17 kernel. The comparison highlights the per-core performance difference between NVIDIA's in-house Olympus cores and ARM Cortex IP, noting these are chips designed for different use cases (desktop/workstation vs. high-end server).

China takes the supercomputer crown without US chips

China's LineShine supercomputer has topped the TOP500 list for the first time since 2017, achieving 2.198 exaflops using an all-CPU design built entirely without Nvidia, AMD, or Intel chips. Housed at the National Supercomputing Center in Shenzhen, it uses custom LX2 processors based on Armv9, runs KylinOS (a Linux variant), and uses a homegrown interconnect called LingQi. The machine is linked to Huawei. While it leads in high-precision scientific computing, it ranks only fourth on AI-style mixed-precision benchmarks, and major US hyperscaler clusters don't even enter the contest. The achievement is framed as a direct response to US export controls, which critics say have inadvertently accelerated China's push for semiconductor self-sufficiency — and exposed a loophole since CPU exports face far looser restrictions than GPUs.