Geekbench AI is a new cross-platform tool designed to evaluate the performance of devices under compute-intensive workloads, with a particular focus on machine learning tasks. It provides full precision, half precision, and quantized scores by measuring the performance of CPUs, GPUs, and NPUs, and includes accuracy assessments. It supports various machine learning frameworks, including ONNX, CoreML, TensorFlow Lite, and OpenVINO. This tool is available on Windows, macOS, Linux, Android, and iOS.

The T-MAC (Table-Lookup-based MAC) technology aims to address the memory and computational constraints of deploying large language models (LLMs) on edge devices. By quantizing model weights into low-bit representations, T-MAC replaces traditional multiplication operations with look-up tables (LUT), significantly enhancing running efficiency on CPUs. This method greatly reduces the memory required for calculations, enabling billion-parameter LLMs to run efficiently on resource-constrained devices for smart upgrades. Compared to existing implementations, T-

In the current wave of AI technology, innovative applications are emerging endlessly, especially in the field of video generation. Recently, an AI tool that transforms highly abstract concepts into video content has garnered widespread attention. These tools can convert popular abstract memes or concepts on the internet into concrete visual presentations, thereby creating content that is both humorous and thought-provoking. For instance, Bilibili UP主 "一朵五颜六色的白云" utilized this AI techno

Fujitsu is developing the next-generation AI, HPC, and data center processor Monaka, which features 150 Armv9 cores. Monaka is expected to utilize 2nm technology and be released in 2027. It is equipped with the SVE2 instruction set, supporting DDR5 memory and PCIe 6.0 interfaces. Monaka employs a 3D chip design for scalability through its separated architecture. Positioned as an AI and data center processor, enhancing energy efficiency is one of its key features.

The ARM CEO stated that they will not design products similar to Nvidia's H100 GPU, believing that Nvidia excels in the AI training field. ARM will continue to focus on CPU technology. Compared to 2016, ARM's business has become more diversified. The revenue proportion from the mobile business has decreased from 60% to 45%, with cloud computing and automotive being the two fastest-growing areas. The IPO will not change ARM's strategy in the Chinese market and will continue to support its partner, Anmou Technology.