炒股就看金麒麟分析师研报，权威，专业，及时，全面，助您挖掘潜力主题机会！

数天前，英伟达发布了NVIDIA DLSS SDK 310.3.0，除了错误修复和稳定性改进外，还明确表示Transformer模型已正式退出beta测试。按照英伟达之前的说法，在Transformer模型基础上不断改进，预计在未来数年里，图像质量会持续提升。

英伟达表示，在NVIDIA DLSS SDK 310.3.0中优化了Transformer模型的显存使用，相比于之前的310.2.0，从1080P到8K分辨率，不同程度减少了使用的显存容量，平均降幅约为20%。这意味着DLSS只需要更少的显存，就能运行超分辨率及光线重建模型。

预览

Transformer模型通过改进的时间稳定性、减少鬼影以及运动中的更高细节来提升图像质量，将进一步提升RTX 20/30/40系列显卡的DLSS性能体验。之前DLSS使用卷积神经网络（CNN）通过分析局部上下文并在连续帧中跟踪这些区域的变化来生成新像素，而新的Transformer模型使用了视觉Transformer变压器，使自注意力机制操作能够评估整个帧中每个像素的相对重要性，并且跨越多个帧。

海量资讯、精准解读，尽在新浪财经APP

Disclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.