Skip to main content

2 docs tagged with "Super"

View all tags

reComputer Super vs Classic Next-Gen AI Performance in Edge Inferencing

This article highlights the superior AI performance of the reComputer Super compared to the reComputer Classic, demonstrating up to a 1.7x boost in AI compute power with the NVIDIA Jetson Orin NX 16GB. Through practical benchmarks in AI text generation using the DeepSeek-R1:7B model with Ollama, and AI video processing using the YOLOv11 object detection model, the results show that the reComputer Super delivers significantly faster inference speeds and higher GPU frequencies. Notably, it processes approximately 2.37 times more images per second than the Classic. The article also explains the importance of evaluating performance based on terminal-reported FPS rather than on-screen FPS, which can be limited by display and system factors.

reComputer Super 与 Classic 的下一代边缘推理 AI 性能对比

本文重点介绍了 reComputer Super 相较于 reComputer Classic 的卓越 AI 性能,使用 NVIDIA Jetson Orin NX 16GB 展现了高达 1.7 倍的 AI 计算能力提升。通过实际基准测试,包括使用 Ollama 的 DeepSeek-R1:7B 模型进行 AI 文本生成,以及使用 YOLOv11 目标检测模型进行 AI 视频处理,结果显示 reComputer Super 提供了显著更快的推理速度和更高的 GPU 频率。尤其是,它每秒处理的图像数量约为 Classic 的 2.37 倍。文章还解释了基于终端报告的 FPS 而非屏幕 FPS 来评估性能的重要性,因为屏幕 FPS 可能受到显示和系统因素的限制。