ARM Ltd. joined a growing roster of processor specialists zeroing in on artificial intelligence and machine learning applications with the introduction of two new processor cores, one emphasizing performance, and the other efficiency.
The chip intellectual property vendor unveiled its high-end Cortex-A75 paired with its “high-efficiency” Cortex A-55 processors during this week’s Computex 2017 event in Taipei, Taiwan. Along with greater efficiency and processing horsepower, the chipmaker is positioning its latest processors as filling the gap in cloud computing by boosting data processing and storage on connected devices.
Along with accelerating AI development, ARM also is advancing its flexible processing approach that incorporates a so-called “big” and “LITTLE” processor core configuration into a single computing cluster. That architecture is based on the assumption that the highest CPU performance is required only about 10 percent of the time. The company also argues that “big” cores can run faster when “little” cores handle low-level workloads.
Based on its DynamIQ multicore architecture technology previewed in March, the Cortex-75 targets emerging AI and machine learning workloads with a single-threaded performance boost of 50 percent, ARM claimed.
“New workloads and their processing requirements are still evolving, so fixed-function dedicated hardware accelerators may not address the newest [machine learning] algorithms,” ARM engineer Stefan Rosinger argued in a blog post. “It makes sense, in that case, to have general CPU capacity….”
Hence, the chipmaker eschews a one-size-fits-all approach by combining a general-purpose processor, “dedicated accelerators” and graphics processing in a system-on-chip as a way of achieving the highest efficiency, Rosinger continued.
The company said it tweaked the Cortex A-75 to deliver a 40 percent boost in infrastructure performance compared to its earlier A-72 processor core for handling machine learning and other complex workloads. The high-end core leaves headroom for emerging workloads, and also targets server and networking applications as ARM seeks for make inroads in x86-dominated datacenters as well as edge devices that would flesh out Internet of Things architectures.
Meanwhile, the Cortex A-55 includes ARM v.8 architecture extensions along with dedicated machine learning instructions. ARM claims an 18-percent performance boost over the previous Cortex version, but the processor’s improved power efficiency points it primarily at IoT edge devices.
Along with processing performance and power efficiency, ARM is betting the rise of AI will heighten requirements for securing data as more personal information is processed and stored on edge devices. “We need to enable faster, more efficient and secure distributed intelligence between computing at the edge of the network and into the cloud,” noted ARM’s Nandan Nayampally.
Meanwhile, ARM said it has so far lined up “more than 10” licensees for both new processors and its DynamIQ framework. Rosinger said he expected initial devices based on the processor cores in early 2018.
ARM was acquired last July by SoftBank Group (TYO: 9984) to expand the Japanese technology conglomerate’s foray into IoT expansion. ARM’s overarching IoT strategy focuses on developing and scaling its Cortex-M 32-bit microcontrollers and a device server that handles connections from IoT devices.