Technology
GPUs
A parallel processing powerhouse: GPUs (Graphics Processing Units) accelerate high-demand workloads like 4K gaming, deep learning AI (e.g., LLMs), and scientific simulation via thousands of specialized cores.
The GPU is a specialized electronic circuit designed for massive parallel processing, fundamentally shifting from its original role in 3D graphics rendering to general-purpose computing (GPGPU). Unlike a CPU's few powerful cores, a GPU employs thousands of smaller, efficient cores (like NVIDIA's CUDA cores) to handle simultaneous calculations, which is critical for modern workloads. This architecture drives performance in high-fidelity gaming (4K/VR), video editing, and especially in AI/Machine Learning: the Tensor Cores, for instance, deliver significant FLOPS for training large language models (LLMs). GPUs are now the core engine for high-performance computing (HPC) and data centers.
Related technologies
Recent Talks & Demos
Showing 1-10 of 10