Arm Introduces TCS23 Platform with Visual Computing Cores and More – News

The TCS ecosystem23

Last week, Arm released Arm Total Compute Solutions (TCS23), an intellectual property (IP) suite from Arm based on the latest ARMv9 architecture. This package offers designers a mix-and-match approach to mobile computing applications through a portfolio of CPU, GPU, system, and software IPs.

The TCS ecosystem23

The TCS ecosystem23. Image used courtesy of Arm

One of the standout features of last year’s TCS22 was the introduction of hardware raytracing on the Immortalis Arm GPU. This year’s announcement boosts the CPU and GPU offerings, primarily through power and efficiency improvements with similar or improved performance.

CPU Clustering for Mobile Computing

Included in the TCS22 are three new lines of Arm Cortex CPUs focused on computing for mobile devices such as smartphones, tablets and laptops. In particular, Cortex processors are designed to enable graphics-intensive applications such as gaming, augmented/virtual reality, machine learning, and on-device security.

Bark-X4

The Cortex-X4 is a fourth-generation high-performance CPU core in Arm’s X-series lineup. Based on the latest ARMv9.2 architecture, the device features a clock speed of 3.4GHz, L1 cache (65KB), L2 cache (0.5–2MB), and L3 cache (32MB). Additionally, the instruction bandwidth has been increased to 10 instructions per cycle. These updates reportedly result in a 15% performance boost over the last generation Cortex CPU.

Performance of the new Cortex CPUs

The Cortex-X4’s 10x performance boost and Cortex-A720’s 4x performance boost make it a powerful contender in the Arm-based high-performance laptop landscape. Image used courtesy of Arm

Cortex-A720

The new 64-bit Cortex-A720 focuses on power efficiency, particularly in applications that need to balance performance and battery life in a big.LITTLE heterogeneous computing configuration. It features L1 I-cache and D-cache (32KB or 64KB), L2 cache (128KB/256KB/512KB), and optional L3 cache (256KB to 32MB). With this device, Arm reports a 20% improvement in energy efficiency and a 4.5% performance improvement over the previous generation.

Cortex-A520

Finally, the 64-bit Cortex-A520 is another updated CPU, focused on efficiency for light, background task loads in a big.LITTLE configuration. The CPU features L1 I-cache and D-cache (32KB or 64KB), optional L2 cache (128KB/192KB/256KB/384KB/512KB), and optional L3 cache (256KB to 32MB). Compared to the last generation, Arm claims an improvement in energy efficiency of 22% and an improvement in performance of 8%.

DynamIQ Shared Unit 120 (DSU-120) supports up to 14 class A architecture CPU cores in a cluster. It provides a shared L3 memory system, security features, control logic, and SoC interface. Arm has unveiled several basic configurations that can support applications ranging from wearables to high-performance laptops, supporting the “mix-and-match” approach.

Schematic of the DynamIQ 120 shared drive

Schematic of the DynamIQ 120 shared drive. Image used courtesy of Arm

New GPUs for various graphics and power needs

TCS23 also includes three fifth generation GPU architectures. Their core technical configurations are nearly identical, with some key differences in the number of cores supported. In some cases, optional ray tracing features are also available. These are mobile GPUs meant to support AAA mobile games, like Genshin Impact or Fortnite.

The first is the Immortalis-G720 targeted at flagship smartphones for gaming and machine learning. It features Deferred Vertex Shading, 2x/4x/8x/16x multi-sampling anti-aliasing, raytracing, and API support for OpenGL ES 1.1/2.0/3.2, Vulkan 1.3, and OpenCL 1.2/2.1/3.0 Full Profile. The GPU is scalable to 10 or more cores and also provides hardware ray tracing capabilities, making it the best performing of the three GPU offerings. Immortalis-G720 is said to have a 15% performance boost, while using 40% less memory bandwidth than the latest generation of Arm GPU architectures.

The Mali-G720 shares the same specs as the Immortalis-G720, though it’s only scalable from six to nine cores, and has optional hardware ray-tracing capabilities. Last but not least is the Mali-G620, which again shares the same base configuration but only scales from one to five cores, providing the lightest GPU option.

Machine learning and safety benefits of TCS23

Arm says the TCS23 enhances machine learning capabilities through a combination of hardware and software.

For CPU-based machine learning applications such as INT8 inference for object detection/classification, real-time recognition, and body position tracking, the company reported a 12% improvement for Cortex-X4, a 9% improvement for the Cortex-A720 and a 13% improvement for the Cortex-A520 over its predecessors. As for the GPU, optimizations in Arm NN and Arm Compute Library provided a 4x improvement in machine learning tasks.

TSC23 also provides some security improvements. The platform supports Android Virtualization Framework for Arm 64-based devices, Pointer Authentication (PAC) and Branch Target Identification (BTI) to eliminate Return Oriented Programming (ROP) and Jump Oriented Programming (JOP) attacks.

#Arm #Introduces #TCS23 #Platform #Visual #Computing #Cores #News

Previous articleThe technology inside Apple’s Vision Pro AR headset: M2 chip, 4K display
Next articleApple’s hands-on approach to artificial intelligence: no bragging rights, just functionality

LEAVE A REPLY

Please enter your comment!
Please enter your name here