几个小时前，NVIDIA CUDA Toolkit 13.1 正式发布，英伟达官方表示：「这是 20 年来最大的一次更新。」 这个自 2006 年 CUDA 平台诞生以来规模最大、最全面的更新包括： NVIDIA CUDA Tile 的发布，这是英伟达基于 tile 的编程模型，可用于抽象化专用硬件，包括张量核心。

编译 | 核子可乐、Tina 今年，英伟达可谓是全力押注，明确表示要确保 Python 成为 CUDA 并行编程框架中的“一等公民”。 多年以来，英伟达为 GPU 开发的 CUDA 软件工具包始终缺少原生 Python 支持，现如今这种情况终于有所转变。在近期的 GTC 大会上，英伟达 ...

In a move that shouldn’t be that surprising, NVIDIA has announced that its popular CUDA platform is being ported to x86. The obvious angle here is that this will give NVIDIA a weapon against OpenCL ...

NVIDIA has announced that it is porting its popular GPU programming architecture to x86. Once the port is complete, developers will be able to choose from two different architectures—OpenCL and ...

Graphics processing units from Nvidia are too hard to program, including with Nvidia's own programming tool, CUDA, according to artificial intelligence research firm OpenAI. The San Francisco-based AI ...

Most notably, the chipmaker announced a compiler source code enabling software developers to add new languages and architecture support to Nvidia’s CUDA parallel programming model. The new ...

在 AI 领域，英伟达开发的 CUDA 是驱动大语言模型（LLM）训练和推理的核心计算引擎。 不过，CUDA 驱动的 LLM 推理面临着手动优化成本高、端到端延迟高等不足，需要进一步优化或者寻找更高效的替代方案。 近日，CMU 助理教授贾志豪（Zhihao Jia）团队创新玩法 ...

Graphics processing units (GPUs) were originally designed to perform the highly parallel computations required for graphics rendering. But over the last couple of years, they’ve proven to be powerful ...

Back in 2000, Ian Buck and a small computer graphics team at Stanford University were watching the steady evolution of computer graphics processors for gaming and thinking about how such devices could ...