![](https://www.techmonitor.ai/wp-content/uploads/sites/29/2019/05/hal-gatewood-484498-unsplash.jpg)
UK-based semiconductor and software designer Arm has launched new CPU and GPU architectures, both aimed solidly at the growing machine learning and gaming sectors.
The Arm Cortex-A77 is the British company’s second CPU in its high-performance family and is built using many of the designs tried and tested in the development of Arm’s Cortex-A76. The Cortex-A77 is a low-power high-performance microarchitecture designed for 10nm, 7nm and potentially 5nm process nodes.
Arm’s latest CPU is designed with the coming generation of smart devices firmly in mind as it aims to facilitate the operation of machine learning (ML) and augmented reality (AR) applications on smart devices, both of which should become more prevalent thanks to moving trends and the rollout of 5G technologies.
Arm noted in a release that: “The CPU is more critical than ever as it handles general-compute tasks, as well as much of the device’s ML compute which must scale beyond today’s limits. The same holds true for more immersive untethered AR/VR applications, and HD gaming on the go.”
Arm CPU Cortex-A77
Essentially Arm has done an extensive redesign and enhancement on the Cortex-A76 architecture to create the A-77. In doing so it claims it has boosted integer performance by 20 percent and increased memory bandwidth improvements by 15 percent.
When the Cortex-A76 was first released in 2018, a key feature was its low-power performance and battery life of over 20 hours. The Cortex-A77 has the same battery life features, but devices will now be capable of extra performance from the CPU within that same power drain window.
With the growing uptake of mobile gaming applications performance improvements at little cost to battery life will be welcomed by users, as these applications are well known to drain batteries at astonishing rates due to their high CPU requirements and intensive graphical displays.
The Cortex-A77 has a 20 percent increase in instructions per cycle (IPC) performance in comparison to its predecessors the A-76.
Stefan Rosinger director of product management notes in an Arm technical blog that with the Cortex-A77: “There is double the branch-prediction bandwidth, various next generation improvements to branch prediction accuracy (which lessens the number of costly branch mispredicts), and increased branch target buffer (BTB) capacity with a 33 percent larger L2 BTB and 4x larger L1 BTB.”
“Meanwhile, the introduction of a Macro-op (Mop) Cache is a fundamental enabler of performance. This feature enables higher fetch bandwidth, lower fetch latency, and dynamic code optimizations to alter the instruction sequence to run more optimally on the downstream core and back-end.”
Arm GPU Mali-G77 Target Next Generation of Devices
Arm has also launched this week the Arm Mali-G77 GPU, which is the company’s first premium GPU based on the Valhall architecture, marking Arm’s move away from the Bifrost architecture. The new Arm GPU design is aiming to improve performance on devices that require GPU’s to run ML and AR applications, as well as high-end gaming.
Andy Craigen director of product marketing at ARM commented in a blog that: “The improved gaming performance of Mali-G77 is linked to the quad texture mapper, which provides four texels/ cycle. This is 2x greater throughput than Mali-G76 and 4x greater than Mali-G72.”
“The quad texture mapper provides improvements across the board of high-fidelity and casual gaming, but has an especially large impact on texture heavy games. In addition, because we increased the compute capability in Mali-G77, we also need to increase the texture capability to keep the machine balanced. This enables Mali-G77 to deliver more performance per square millimetre than ever before.”
The Arm GPU Valhall architecture brings new features such as;
- A new superscalar engine, which delivers another leap in energy efficiency and performance density
- A simplified scalar ISA with a new instruction set that is more compiler friendly
- New dynamic scheduling of instructions
- Reworked datastructures better aligned to modern APIs, such as Vulkan.
Arm is firmly aiming this new design at OEMs working to create smart devices that can handle complex AR and ML workloads, while also being capable of handling the increases of bandwidth heavy applications that 5G is about to unleash onto the market.