产品精选
IBM 通过 PyTorch 现有抽象层集成 Spyre 加速器
精选理由
IBM 把自家 Spyre 芯片接进了 PyTorch，靠的是设备、内存、编译这些现成抽象层，想了解框架如何兼容多硬件的可以看看这篇官方博客。
PyTorch Foundation 发布博客，介绍来自 IBM 的贡献者如何将 Spyre 硬件接入 PyTorch。集成复用了 PyTorch 现有的设备、流、内存管理和编译抽象，没有为 Spyre 单独建框架。结果是开发者体验更统一，Spyre 运行时更稳定，PyTorch 生态得以覆盖更多硬件。
原文 · PyTorch
How can a single framework support an increasingly diverse range of AI hardware?
In a new PyTorch Foundation blog, contributors from @IBM share how Spyre was integrated using existing PyTorch abstractions for devices, streams, memory management, and compilation.
The result is a more seamless developer experience, a more robust Spyre runtime, and a PyTorch ecosystem that can support a broader range of hardware without requiring a separate framework for each device.
Learn more here: https://t.co/rmh4ILOL8v