The Cortex-A75 and Cortex-A55 tout more performance and efficiency over their predecessors
ARM has revealed a suite of next generation mobile processors and graphics accelerator designs, targeted at enabling artificial intelligence (AI) on smartphones and Internet of Things (IoT) devices.
First out of the blocks is ARM’s latest flagship chip architecture, the Cortex-A75, which ARM claims delivers a 22 percent boost in performance over its predecessor the Cortex-A73 and offers better multi-core abilities compared to previous designs.
The Cortex-A75 is aimed at both smartphones an interestingly laptops, traditionally the domain of Intel chips, which would suggest that ARM’s flagship chips are being geared up for supporting efforts to bring Microsoft’s Windows onto ARM’s instruction set.
ARM also introduced the Cortex-A55, a chip design targeted at power efficiency, offering a 15 percent boost compared to its predecessor.
On the graphics side, ARM’s new Mali-G72 graphics processing unit (GPU) also comes offering a 25 percent hike in performance over the older Mali-G71.
ARM-ed for machine learning
But more interesting is the introduction of Dynamiq into the chip designs, the name given to ARM new range of flexible designs, which marks an evolution of the big.LITTLE architecture.
Multicore chipsets using the big.LITTLE design, match low-powered cores for handling undemanding tasks and ensuring battery life is preserved, with higher speed cores aimed a taking care of more demanding tasks. Usually these cores are arranged in binary layouts; two or for low-powered cored matched with the same number of high-performance cores.
Dynamiq mixes this up a little by enabling chipmakers to mix and match more core combinations to suit the tasks and compute activities required from the processors; for example an eight core chipset could be made up of one powerful core like the Cortex-A75 and seven less power hungry Cortex-A55 processors.
With this flexibility also comes the capability to better run machine learning algorithms on mobile chips and IoT devices, rather than relying on connectivity to cloud based smart systems.
“Unfortunately, a cloud-centric approach is not an optimal long-term solution if we want to make the life-changing potential of AI ubiquitous and closer to the user for real-time inference and greater privacy,” said Nandan Nayampally, ARM’s vice president of marketing and strategy.
“Enabling secure and ubiquitous AI is a fundamental guiding design principle for ARM considering our technologies currently reach 70 percent of the global population. As such, ARM has a responsibility to rearchitect the compute experience for AI and other human-like compute experiences. To do this, we need to enable faster, more efficient and secure distributed intelligence between computing at the edge of the network and into the cloud.”
As such, the championed power, efficiency and flexibility of the new ARM designs are aimed at taking AI and machine learning capabilities to the devices on the edge of IoT networks.
This approach looks to allow for smart and AI-based technology to be used in a rapid and seamless fashion; for example, the use of image recognition in autonomous cars or speedy natural language cognition from the virtual assistants that are now commonplace on higher-end smartphones.
ARM’s designs are due to make their debut in chips from the likes of Qualcomm in early 2018 and as such we can expect to see successors to flagship chipsets like the Snapdragon 835.
What do you know about the Internet of Things? Take our quiz!