赞
踩
The Cortex A53 uses the ARMv8-A architecture to support 32-bit ARMv7 code and 64-bit AArch64 execution state. The Cortex-A53 processor architecture was released in 2012 designed for reduced power consumption and improved energy efficiency. It offers better performance than the Cortex-A7 processor and can be used as a stand-alone major application processor, or with the Cortex-A57 processor to form a big.LITTLE configuration. The Cortex-A53 offers higher performance than the Cortex-A9 at the same frequency.
The Cortex-A15 was first released in 2010 and is based on the 32-bit ARMv7-A architecture. The Cortex-A15 has 32 KB of Instruction and Data Cache respectively. The Cortex-A15 has a pipeline depth of 15 whereas the Cortex-A53 has only 8, making the A15 faster. The ARM Cortex A15 was the next generation processor that ended up in a huge number of devices in 2013. The 249$ Chromebook and the Nexus 10 were both powered by Cortex A15-based SoCs.
The speed of a given CPU depends on many factors, such as the type of instructions being executed, the execution order etc. The MIPS (million instructions per second) value can be useful when comparing performance between processors made with similar architecture, However, MIPS doesn’t make sense when processors are based on different architecture. For this reason, DMIPS / MHz (Dhrystone MIPS) is considered whilst estimating the performance of CPUs, wherein DMIPS result is divided by CPU frequency to enable easy comparison of CPUs running at different clock rates.
DMIPS is a sort of a simple benchmark without I/O tests. The Dhrystone benchmark provides a measure of integer performance (no floating point instructions). The DMIPS/Mhz (Dhrystone MIPS/Megahertz) values are sourced from multiple websites. The DMIPS value for Cortex-A75 and Cortex-A55 are estimated using the performance increase as claimed by ARM. As claimed, “Cortex-A55 delivers significantly higher performance than the Cortex-A53 across the board, including integer (+18%)” and “Cortex-A75 processor provides a significant boost in single-thread performance using a fully out-of-order, variable-length, and symmetrical three-way superscalar pipeline. With over 20% more integer core performance”.
Copyright © 2003-2013 www.wpsshop.cn 版权所有,并保留所有权利。