ARM released a new generation of high-performance CPU architecture design – Cortex A76, and a new generation of GPU architecture design – Mali-G76. ARM stated that the performance of the A76 architecture, such as performance, power control, and other performance is significantly increased than the previous generation A75 core.
ARM Cortex A76 CPU Core
The Cortex A76 is designed for 7nm process nodes at frequencies up to 3GHz, while the comparable Cortex-A73 and Cortex-A75 are 2.45GHz for the 14/16nm node and 2.8GHz for the 10nm node, respectively. Cortex-A76 has advantages in both architecture and frequency.
The Cortex-A76 will still support ARM’s DynamlQ size core technology, which is also up to 8 cores, but is more flexible because the Cortex-A76 can be paired with different generations of CPUs, such as the Cortex-A76 and the Cortex-A55 core.
In addition, according to officials, the Cortex A76 can use not only the 7nm process but also the TSMC’s 16nm FFC process, which is one of the most cost-effective options.
Compared with the previous generation 10nm A75 core, the new generation A76 architecture has achieved a 35% improvement in performance, a 40% increase in energy efficiency, and a 4x improvement in machine learning capabilities.
After Cortex A75, the ARM Cortex A76 CPU is the second high-performance processor core based on DynamIQ technology. ARM also stated that if a new architecture CPU is used on a notebook, it can achieve twice the performance of the current version of the CPU.
ARM Cortex A76 Highlights:
Architecture – Armv8-A (Harvard) with Armv8.1, Armv8.2, Armv8.3 (LDAPR instructions only), cryptography and RAS extensions
ISA support – A64; A32 and T32 (at the EL0 only)
Pipeline – Out-of-order
NEON / Floating Point Unit
Optional Cryptography Unit
Up to four CPUs in cluster
Physical addressing (PA) – 40-bit
Memory system and external interfaces
64KB L1 I-Cache / D-Cache
256KB to 512KB L2 Cache
Optional 512KB to 4MB L3 cache
ECC Support, LPAE
Bus interfaces – AMBA ACE or CHI
Optional Peripheral Port
ARM Mali-G76 GPU Core
In addition to the Cortex-A76, ARM has also released the Mali-G76 GPU core. The architecture is also the beginning of the G7X series Bifrost. Based on this, it has fine-tuned and is mainly aimed at high-performance mobile gaming platforms.
In terms of architecture, the number of cores of the G76 has been reduced from the previous 32 to 20, but the number of execution channels in each engine has doubled, and the total number of execution units has increased from 384 in the G72 to 480.
As for performance improvement, the official said that compared to the Mali-G72 GPU core, the performance density of the Mali-G76 GPU core is increased by 30%, energy efficiency is improved by 30%, and ML deep learning performance is improved by 2.7 times.
Mali-G76 GPU Specifications:
Anti-Aliasing – 4x MSAA, 8x MSAA, 16x MSAA
API Support – OpenGL ES 1.1, 2.0, 3.1, 3.2, Vulkan 1.1, OpenCL 1.1, 1.2, 2.0 Full Profile
Bus Interface – AMBA 4, ACE-LITE
L2 Cache – 512KB to 4MB
Scalability – 4 to 20 Cores
Adaptive Scalable Texture Compression (ASTC) – Low Dynamic Range (LDR) and High Dynamic Range (HDR), supports both 2D and 3D images.
Arm Frame Buffer Compression (AFBC) – Version 1.2; 4×4 pixel block size
The GPU will be used in “premium mobile”, virtual reality, machine learning, and automotive applications. For more information, check out the blogpost and product page on Arm’s website.
ARM Mali-V76 VPU Core
In addition to the A76 and G76 , ARM also announced a video decoder called Mali-V76 , which will provide 2-8 core designs. Mali-V76 is the latest video processing unit (VPU) from ARM with support for 8K video decoding @ 60 fps, and also suitable for video walls with 2×2 4K UHD videos, or 4×4 1080p HD videos.
Mali-V76 VPU Highlights:
Multi-standard video processor
10/8-bit HEVC, VP9, VP8, H.264, AVS+/AVS and legacy
Simultaneous encode and decode
Scalable 2-8 cores (8K60D/8K30E)
Mali-V76 is an evolution of Mali-V61 video processor with twice the decode performance, a 40% smaller area for 4K120 performance, 25% additional bitrate saving, twice the bus fabric latency tolerance, and additional support for 10-bit H.264 codec and 8-bit AVS+/AVS decode.
SoC (System-on-Chips) based on the new Cortex A76 CPU and GPU IP will provide “laptop-class” performance, and the company also announced Arm Mali-V76 VPU with support for 8K video decoding and encoding. ARM is currently working hard to open up the Windows PC market.
Prior to this, foreign media has also revealed a new generation of Qualcomm Snapdragon 1000 SoC in the making to power Windows 10 ARM notebooks. The Always-Connected PC products from Microsoft and its partners will give ARM a huge opportunity. One of the important features of the Always-Connected PC is long battery life. If ARM can succeed on Windows 10 on ARM, it will greatly affect Intel’s market position.
DC works as a technology analyst in a research and advisory firm specializing in automotive, energy, electronics, and semiconductors domain. When he's not working, he loves gaming on his modest pc build, loves sci-fi films, and big pop culture geek!