Generated by GPT-5-mini| ARM Cortex-A57 | |
|---|---|
| Name | ARM Cortex-A57 |
| Designer | ARM Holdings |
| Architecture | ARMv8-A |
| First release | 2012 |
| Cores | 1–8 per cluster |
| L1 cache | configurable |
| L2 cache | configurable |
| L3 cache | implementation-dependent |
| Process | 28 nm, 20 nm, 16 nm, 14 nm, 10 nm variants |
ARM Cortex-A57 is a 64-bit processor core designed by ARM Holdings as part of the ARMv8-A family, introduced during the early 2010s to bring high-performance computing to mobile, embedded, and server markets. It was announced alongside contemporaries and successors in ARM's roadmap and targeted partners such as Qualcomm, Samsung Electronics, NVIDIA, Broadcom, and AppliedMicro. The core sought to balance single-thread throughput, multicore scalability, and compatibility with ecosystems around Linux (kernel), Android (operating system), and proprietary real-time operating systems from companies like Green Hills Software and Wind River Systems.
The Cortex-A57 implements the 64-bit ARMv8-A instruction set architecture, supporting AArch64 and AArch32 states and extensions used by vendors such as Cavium, Marvell Technology Group and Huawei. Its microarchitecture uses an out-of-order pipeline inspired by earlier ARM cores and influenced by designs from partners including Texas Instruments, STMicroelectronics, and MediaTek. The design incorporates multiple levels of caches that integrators from Xilinx and Intel-fab partners could scale, and includes support for virtualization features embraced by VMware, Red Hat, and Canonical (company). Security extensions trace to work by organizations like ARM TrustZone and have been deployed in systems by Apple Inc. and Lenovo.
The Cortex-A57 features a deep pipeline with branch prediction, superscalar issue, and a wide execution engine that benefited products from Qualcomm Snapdragon and Samsung Exynos families. Performance characteristics were evaluated by teams at SPEC (organization), Phoronix, and academic groups at MIT, Stanford University, and University of Cambridge. Microarchitectural elements such as load/store units, integer and floating-point pipelines, and NEON SIMD units were comparable to contemporaneous cores used by NVIDIA Tegra platforms and influenced designs at ARM Cortex-A72 and Cortex-A53 pairings in big.LITTLE configurations used by Sony, HTC, and LG Electronics. Benchmark reports appeared in outlets including AnandTech, Ars Technica, and Tom's Hardware.
Multiple silicon vendors licensed the Cortex-A57, including Samsung Electronics, Qualcomm, NVIDIA, Broadcom, Marvell Technology Group, AppliedMicro, HiSilicon, MediaTek, Texas Instruments, Fujitsu, Hitachi and NXP Semiconductors. Implementations were fabricated at foundries such as TSMC, GlobalFoundries, Samsung Foundry, and Intel Foundry Services nodes spanning 28 nm to 10 nm processes. OEMs and ODMs like Foxconn, Quanta Computer, Pegatron, and Compal integrated these SoCs into devices released by Dell Technologies, HP Inc., Lenovo, Asus, and Acer. Networking and storage appliances used A57 silicon in products by Cisco Systems, NetApp, and Juniper Networks engineers.
Cortex-A57 cores were commonly combined into heterogeneous SoCs pairing with low-power cores such as ARM Cortex-A53 in big.LITTLE clusters adopted by Samsung Galaxy series, Google Nexus devices, and Chromebooks supported by Google LLC and Chromium OS developers. SoC integrators included modem and GPU vendors like ARM Mali, Imagination Technologies, and Adreno teams within Qualcomm. Platforms ranged from mobile smartphones and tablets to microservers used by Amazon Web Services partners and embedded gateways from Siemens and Schneider Electric. Reference platforms and development boards were provided by companies like NXP, BeagleBoard, and Raspberry Pi Foundation-adjacent projects.
Software ecosystems for Cortex-A57 involved contributions from Linux Foundation, Canonical (company), Red Hat, SUSE, and commercial BSD projects like FreeBSD. Major operating systems ported to A57 silicon included Android (operating system), Chrome OS, Ubuntu, Debian, Fedora, and vendors' real-time systems such as QNX from BlackBerry Limited. Compiler and toolchain support came from GCC (GNU Compiler Collection), LLVM Project, and commercial compilers by ARM Ltd. and Mentor Graphics. Virtualization stacks and cloud orchestration used hypervisors from KVM (kernel-based virtual machine), Xen Project, and containers orchestrated with Docker and Kubernetes on A57-based servers.
Power and thermal behavior of Cortex-A57 silicon were critical for deployment in smartphones by Samsung Electronics and tablets by Apple Inc. competitors, and in microservers by Calxeda and AppliedMicro. Techniques such as dynamic voltage and frequency scaling (DVFS), big.LITTLE switching, and heterogeneous multiprocessing were integrated with power management IP from ARM Power Management partners and firmware stacks from Coreboot contributors. Thermal solutions involved collaboration with OEM thermal teams at Dell Technologies, HP Inc., and cooling suppliers like Cooler Master for higher-TDP deployments.
The Cortex-A57 received attention in reviews by AnandTech, The Verge, Wired, and CNET for its step-up in 64-bit performance versus earlier ARM cores and its role enabling ARM servers championed by Calxeda and later by Cavium and Ampere Computing. Benchmarks from SPEC (organization), Geekbench, and Linpack highlighted strengths in integer and floating-point workloads, while critiques pointed to thermals and power in smartphone form factors reported by GSM Arena and PhoneArena. The architecture influenced subsequent ARM cores including the ARM Cortex-A72 and later ARMv8-A derivatives used by Apple Silicon-adjacent designs and cloud vendors such as Microsoft Azure and Google Cloud Platform. The A57's ecosystem legacy persists in the software stacks of mainstream Linux distributions and virtualization platforms maintained by Red Hat and Canonical (company).