Imagination AI solution leads performance 2 times closest to competitors

**Introduction** When it comes to AI acceleration, the expert knows what's best. Imagination Technologies' latest neural network accelerator, PowerVR 2NX NNA, is now delivering performance that’s twice as close to the competition as any other solution—whether it's based on DSP + hardware, VPU + hardware, or a dedicated neural network DSP. This makes it an ideal choice for HiSilicon’s next-generation Kirin chips, promising significant improvements in AI processing power and efficiency. Artificial intelligence has become a strategic priority for nations around the world. As AI permeates key equipment across industries, it's becoming the core differentiator for smart devices. For example, Huawei has already outperformed Apple’s iPhone 8 using AI technology. The neural network processing unit that mimics human brain computing models is set to be the top choice for future machine learning and deep learning applications. Chris Longstaff, Senior Director of Product and Technology Marketing at Imagination, said: “Considering power, bandwidth, performance, reliability, security, and latency, the next generation of AI devices will not rely solely on software or cloud computing. Hardware-based solutions for machine learning and deep learning are essential. Neural networks will soon become the key processing unit in AI SoCs, just like CPUs and GPUs.” [Image: A visual representation of the PowerVR 2NX NNA architecture] Neural networks, such as CNNs, RNNs, and LSTMs, are driving major advancements in AI industries. Imagination, a leading GPU IP provider, recently launched a complete, standalone hardware IP neural network accelerator in Shenzhen. This new solution leverages the PowerVR architecture to deliver industry-leading performance, power, and area efficiency. Chris Longstaff added: “The PowerVR 2NX NNA offers performance, power, and area efficiency that is two times closer to the competition than any other solution, regardless of whether the competitor uses DSP + hardware, GPU + hardware, or a neural network DSP.” Today, companies developing SoCs for mobile, surveillance, automotive, and consumer systems can integrate high-performance computing with minimal power consumption and chip area using the PowerVR 2NX NNA. For instance, modern drones flying at over 150 mph require real-time object detection and avoidance. Without hardware acceleration, collision avoidance would need a 10–15 meter range. With PowerVR 2NX NNA, this range can be reduced to less than 1 meter, allowing the drone to run multiple networks for object recognition and tracking simultaneously. Another example: sorting 1,000 photos on a mobile device using a high-end CPU takes 60 seconds, but with PowerVR 2NX NNA, it only takes 2 seconds. NNA is expected to become as essential as CPUs and GPUs. It has wide-ranging applications—from AR/VR headsets (feature detection, eye tracking) to automotive safety (pedestrian detection, driver alertness monitoring), intelligent surveillance (facial recognition, crowd behavior analysis), and more. David Harold, Vice President of Marketing at Imagination, said: “PowerVR 2NX is specifically designed for mobile and Android platforms. It's the first dedicated neural network hardware solution with support for flexible bit widths from 16-bit to 4-bit and low bandwidth requirements.” According to the Embedded Vision Developer Survey, 79% of respondents have used or plan to use neural networks in their products. Imagination’s customers have already deployed NN-based systems in security, mobile, automotive, and set-top box markets. Jeff Bier, founder of the Embedded Vision Alliance, stated: “Dedicated processors like PowerVR 2NX NNA are crucial for deploying powerful algorithms in various applications. They provide the performance needed while keeping cost and power under control.” With growing demand for neural networks, solutions like PowerVR 2NX NNA—offering 8x performance density over pure DSP—are set to redefine AI processing. They reduce power and cost while optimizing bandwidth usage, making them ideal for future AI SoCs. [Image: A comparison of PowerVR 2NX NNA with competing solutions] **PowerVR 2NX NNA – The Most Efficient Solution** Designed from scratch, PowerVR 2NX NNA provides: - **Industry’s highest unit milliwatt inference (mW) IP core** for ultra-low power consumption - **Industry’s highest in-plane/inference (mm²) IP core** for the most cost-effective solution - **Lowest bandwidth solution** with support for flexible bit widths, including 4-bit mode - **High-performance single-core design** with 2048 MAC per cycle, scalable with multi-core setups - **Comprehensive system-level neural network solution** tailored for hardware architecture Chris Longstaff added: “As neural network acceleration becomes standard in future SoCs, PowerVR 2NX NNA sets a new benchmark. It delivers superior performance, power efficiency, and area utilization, helping developers deploy neural networks quickly and effectively.” **Neural Network Application Market Is Everywhere** PowerVR 2NX NNA is designed for inference engines that support multiple markets. Its highly scalable architecture makes it suitable for a wide range of future applications. **Mobile:** With the rise of TensorFlow Lite and Caffe2Go, we're seeing a surge in smartphone AI apps. PowerVR 2NX is the only solution that meets all mobile needs—low power, small area, MMU support, and Android compatibility. **Intelligent Surveillance:** As surveillance cameras grow, so does the need for on-device AI processing. PowerVR 2NX NNA provides efficient, stand-alone acceleration for edge-based video analytics. **Automotive:** From driver monitoring to road sign detection, PowerVR 2NX NNA supports the expanding needs of autonomous vehicles and smart transportation systems. **Home Entertainment:** Set-top boxes and TVs are adopting AI for personalized experiences. PowerVR 2NX NNA helps achieve these features with low bandwidth and cost. **Simplify the Work of Developers** Imagination provides tools that make it easy for developers to enable and run their neural networks. From mapping tools to sample networks, everything is included to optimize performance and accuracy. Imagination also offers a universal DNN API that works across CPU, GPU, and NNA, making prototyping simple and efficient. The PowerVR 2NX NNA is now available for licensing. **About Imagination Technologies** Imagination is a global leader in technology, powering billions of devices worldwide. Its comprehensive portfolio includes silicon IP for SoCs used in mobile, consumer, and embedded electronics. With innovative multimedia, processor, and connectivity technologies, Imagination enables its customers to create differentiated products and bring them to market faster. Its authorized customers include leading semiconductor manufacturers and OEMs behind some of the world’s most iconic devices.

USB 3.1 Cable

UCOAX Custom Made USB 3.1 Type C SuperSpeed Cables combine style, quality, performance and value to give a great deal on a great USB cable.

SuperSpeed USB 3.1 Type C is 3rd revision of the ubiquitous USB (Universal Serial Bus) Standard. Clocking in at speeds up to 10Gbit/s, USB 3.1 Type C is a vast improvement over the USB 2.0 speed of 480 Mbit/s while being completely backwards compatible with USB 2.0.

UCOAX USB 3.1 C Male to C Male Cable

Up to 10Gbit/sec
PC and Mac Compatible
Ultra-flexible jacket; Molded strain relief
Foil and braid shield to guarantee an interference free connection
EMI/RFI interference: Metal connector shield to meet FCC requirements
RoHS Compliant

Usb 3.1 Cable,Type C Superspeed Cable,Usb Micro B Superspeed To Usb C,Super Speed Usb C Cable,Fast Charging USB Cable

UCOAX , https://www.ucoax.com