Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

textbook 3 and 4 #153

Open
xiaowuhu opened this issue Jun 10, 2022 · 4 comments
Open

textbook 3 and 4 #153

xiaowuhu opened this issue Jun 10, 2022 · 4 comments
Assignees

Comments

@xiaowuhu
Copy link
Collaborator

  • 个人感觉第3章和第4章可以合并(从内容连贯性上看)。另外一个理由就是这两章内容都比较少。
@YanjieGao
Copy link
Contributor

Ying and Jilong还准备加内容

@YanjieGao
Copy link
Contributor

@lcy-seso
@jlxue

@lcy-seso
Copy link
Contributor

从目前merge的内容看,可能可以把第4章移动到第三章之前,或者,成为第三章的前一两节。

另外,在写第三章的时候,也觉得目前写出来的内容较短。究其原因,图优化在第5章,分布式计算在第6章,这两部分原先也是框架设计的一部分,目前被分散到其他章节。于是框架设计,就剩下运行时系统里面的内存管理和任务调度。

@YanjieGao
Copy link
Contributor

YanjieGao commented Jun 14, 2022

体系结构挪框架前较为合理,如果第4章不准备扩充内容可以考虑合并,但是我感觉体系结构很重要需要扩充一些内容(NN的硬件也是重要且高速发展的领域)。目前在第8章推理芯片有讲一些来龙去脉,1.5略有介绍,但是没涉及目前主流NV training GPU这块还有system on wafer等新架构,我感觉有几点趋势还需要在体系结构再加强说明。

1 训练芯片的演进脉络
2 NPU几大流派
3 体系结构 for DL的reseach方法论
4 NN芯片的算法硬件化特定加速思路,例如稀疏,量化等做进硬件
5 high level的一些多方向趋势总结

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

4 participants