龙芯发布自研通用 GPU 加速计算平台首个软件版本:支持 OpenCL 3.0、CUDA 及 AI 推理
AI Digest
通用GPU加速计算平台OpenCL 3.0CUDAAI推理引擎算子库优化
龙芯发布首款通用GPU加速计算平台软件版本,支持OpenCL 3.0、CUDA及AI推理,提供完整软件栈和开发工具链。
Loongson launches its first general-purpose GPU acceleration platform with OpenCL 3.0, CUDA, and AI inference support, offering a complete software stack and development tools.
Key points
- 平台支持OpenCL 3.0、CUDA编程模型及AI模型推理 Supports OpenCL 3.0, CUDA programming models, and AI inference
- 集成驱动、编译器、运行时环境等完整软件栈组件 Integrates drivers, compilers, runtime environments, etc.
- 优化BLAS/DNN算子库提升硬件资源利用率 Optimized BLAS/DNN libraries for hardware efficiency
- LacInfer引擎实现ONNX模型直接部署 LacInfer enables direct ONNX model deployment
- 兼容龙芯全系列芯片及Linux多内核系统 Compatible with all Loongson chips and Linux kernels
Takeaway: 国产GPU生态迈出关键一步,降低AI开发适配成本。 / Marks progress in domestic GPU ecosystem, reducing AI development adaptation costs.
Why it matters 展现国产GPU生态建设成果,为开发者提供跨平台算力解决方案。
View original ↗ Back to hot list
This page is an aggregated digest from ithome; content and hot-score data come from public sources. Copyright belongs to the original authors. We link to originals with nofollow and never republish full text.